From Asymmetry to Equilibrium: How Government Regulation Drives Sustainable Digital Asset Management on Media Platforms in China

Abstract

The rapid digitalization of the media and publishing industry has deepened systemic asymmetries in resources, power, and institutional rights. These asymmetries create fundamental barriers to the economic–institutional sustainability of digital content dissemination. Existing governance frameworks have not yet comprehensively addressed the resulting competitive and informational imbalances. Adopting China’s publishing and media industry as a focal case, this study draws on symmetry theory to develop an integrated analytical framework. It reconceptualizes government regulation as a multi-dimensional governance mechanism operating across three dimensions: resource allocation, technological innovation, and rights protection. We test this framework empirically using Xinbang Index data covering the top 10 publishing and media enterprises from 24 January 2025 to 7 December 2025. Multiple regression analysis and Spearman rank correlation are applied to assess each dimension’s differential impact on content dissemination efficiency. The results yield four key findings. First, all three regulatory dimensions contribute positively to content dissemination efficiency. Second, technological innovation is the most potent symmetry-restoring lever, exerting a statistically robust direct effect on dissemination outcomes. Third, resource allocation provides a necessary foundational contribution, while rights protection operates conditionally—its effect is fully realized only alongside adequate technological and resource inputs. Fourth, an integrated multivariate regression confirms the cross-dimensional hierarchy: the standardized Beta coefficient for technological innovation (β = 0.394) exceeds those for rights protection (β = 0.294) and resource allocation (β = 0.125). No single regulatory instrument is sufficient to achieve dynamic equilibrium. A synergistic, technology-centered combination of all three dimensions is required. This study proposes a tripartite symmetry-based governance strategy for media platform ecosystems. The symmetry framework developed here offers an analytical template for diagnosing analogous asymmetries in other platform-dependent sectors. Empirical validation beyond the Chinese publishing and media context is recommended as a priority for future research.

1. Introduction

The new economic growth theory holds that optimal resource allocation and balanced coordination are the primary engines of value creation [1]. In the digital economy, this balanced state can be described as systematic symmetry—a condition where resources, power, and opportunities are equitably distributed across platform actors. Yet the current global digital transformation has intensified profound asymmetries across environmental, economic, and social dimensions [2]. The publishing and media sector illustrates these dynamics clearly. Traditional print and broadcast formats have given way to digital content platforms, streaming services, and algorithmic distribution ecosystems [3,4]. Digital asset management—covering digital copyrights, user data, and algorithmic models—has become the central basis for competitive advantage. However, this transition has neither been smooth nor symmetric. Three distinct asymmetry patterns have emerged. Environmentally, large-scale digital infrastructure generates surging energy demands, revealing a structural mismatch between digital growth and planetary carrying capacity [5]. Economically, algorithmic and platform power has concentrated, pushing markets toward monopolistic structures—a tendency documented by antitrust investigations in China and elsewhere [6]. Socially, barriers to information access and algorithmic bias have deepened the digital divide [7]. These asymmetries constitute fundamental obstacles to the sustainability of digital content dissemination. The environmental dimension lies outside the empirical scope of this study (see Section 2.6).

Government regulation is not a peripheral corrective mechanism. It is a central governance force capable of restructuring platform ecosystems toward dynamic equilibrium. Prior scholarship has examined digital transformation [8], government regulation [9], and sustainable development [10] largely in isolation. A unified framework linking these three domains has not yet been established. Without such a framework, regulatory interventions risk addressing individual symptoms rather than the systemic imbalances that generate them. This study poses the following research question: How do the three regulatory dimensions of resource allocation, technological innovation, and rights protection affect content dissemination efficiency in the publishing and media industry? More specifically, how does their combined operation move the industry toward a more symmetrical distribution of dissemination capacity?

To answer this question, this study pursues three objectives, with China’s publishing and media industry serving as the primary empirical context—a setting chosen for its distinctive combination of state-led platform governance and rapidly concentrated digital market structure. First, it constructs an integrated analytical framework that reconceptualizes each regulatory dimension as a mechanism for correcting a distinct form of systemic asymmetry. Second, it applies multiple regression analysis and Spearman rank correlation to estimate how each dimension—and their combination—influences content dissemination efficiency across Chinese publishing and media enterprises. Third, it derives evidence-based implications grounded in the empirical results.

This study makes three contributions. First, it integrates symmetry theory with a multi-dimensional sustainability framework and applies this combined lens to digital asset governance on media platforms. This combination offers a novel vocabulary for diagnosing systemic imbalances in digital platform ecosystems. Second, it constructs and empirically validates a framework in which government regulation operates as a synergistic mechanism across three coordinated dimensions. This bridges regulatory design, platform-level asset management, and sustainability outcomes in a single testable structure. Third, the empirical evidence for a multi-dimensional, technology-centered approach advances understanding of how structural symmetry is achieved through coordinated rather than single-instrument intervention. Beyond its immediate relevance to the Chinese publishing and media industry, the framework provides an analytical template for examining analogous asymmetries in other platform-dependent sectors. Future research can deploy this three-dimensional structure in contexts such as healthcare information systems or educational technology platforms. Such work would test whether the directional patterns identified here generalize across different institutional and market contexts.

The remainder of this paper is organized as follows. Section 2 and Section 3 develop the theoretical framework, delineating the core dimensions of government regulation, digital asset management, and symmetry theory, and deriving the research hypotheses. Section 4 presents the empirical methodology and data. It reports and interprets the regression and Spearman rank-correlation results for each regulatory dimension. Section 5 discusses the findings in relation to the existing literature and derives theoretical and practical implications. Section 6 concludes with a tripartite symmetrical governance strategy, policy recommendations, and directions for future research.

2. Literature Review

2.1. Asymmetry in Digital Ecosystems: Theoretical Roots of Government Regulation

Government regulation is fundamentally rooted in correcting market asymmetries. Classical regulatory theory views regulation as a response to market failure that protects public interests, with information asymmetry and power asymmetry as primary drivers [9,11]. In the digital economy, these asymmetries have been sharply amplified. Sun Jin (2020) argues that regulation should intervene in market actors according to rules to preserve a healthy competitive order [12]. This rule-based interventionist position resonates with symmetry theory in a specific way: the function of regulation, viewed through a symmetry lens, is precisely to correct the uneven distribution of market power that unconstrained platform competition tends to produce—Sun Jin’s “healthy competitive order” is, in effect, an economic analog of structural symmetry in the market network. However, the emergence of the digital platform economy has produced a new, intensified form of asymmetry: platforms accumulate dominant market power through data, algorithms, and network effects, producing a structural imbalance in which “the big bully the small” [6].

Scholars disagree on the proper role of regulation. Conservative theorists trust the market’s self-correcting mechanisms, while more interventionist scholars argue that a fair competitive environment must be actively constructed through “regulated competition” [13]. This debate is especially salient in digital governance, where regulation is no longer only remedial but is expected to proactively shape the conditions for symmetric competition. Li Dongxu et al. (2025) show that effective antitrust enforcement does not stifle innovation; instead, it prevents abusive power at its source and preserves comparable development opportunities for firms of different sizes, directly addressing competitive asymmetries [6]. From a science and technology studies standpoint, Callon’s (1984) generalized symmetry principle—treating human actors, institutions, and nonhuman entities (technologies, standards, algorithms) on equal interpretive footing [14]—provides this study with its analytical license to treat technology and rights protection as regulatory dimensions on par with resources, rather than as secondary support functions. Without Callon’s generalized symmetry principle, one might collapse technology into a neutral tool and rights into a derivative of resources, missing the ways in which algorithmic architectures and institutional rules themselves produce distinct forms of asymmetry that require their own corrective mechanisms. This is precisely why our three-dimensional framework assigns technology and rights protection the status of independent governance dimensions. Accordingly, this study starts from the premise that a central regulatory function is to correct endogenous asymmetries of power and opportunity in digital markets through institutional design—a function that operates most effectively simultaneously across resource, technological, and institutional rights dimensions in coordination.

2.2. The Sustainable Dilemma of Asset Management: The Asymmetry of Resources and Rights

Research on asset management in the publishing and media industry also highlights a pronounced asymmetry between resources and rights [15]. From an asset definition standpoint, Sun&Yin emphasizes that assets are property interests that yield economic benefits [16], whereas Wu Jieming et al. describe assets as an assemblage of resources, holdings, and products [17]. Together, these views point to the central issue in asset management: symmetry in value distribution. During digital transformation, however, asset management becomes tightly coupled with rights protection, effectively evolving into a rights-management process [17,18].

The operation of digital assets currently faces multiple asymmetries. First, resources are unevenly allocated: large platform firms markedly outperform small and medium-sized content creators in accessing, controlling, and monetizing digital resources. Second, rights protection is asymmetric [19]. As Wu Jing (2015) observed, asset management seeks to preserve and grow value, but in the digital environment the high cost of enforcing rights and the ease of infringement create a sharp mismatch between rights and responsibilities [17]. Similarly, Zou Hanqing (2019) highlights an asymmetry of capability in media databases [20], where firms’ internal technical capacity does not align with external market opportunities. Sustainable asset management is therefore most effective on reducing these asymmetries to foster a more balanced distribution of economic and social value.

2.3. The Dual Role of Digital Technology: Empowering and Exacerbating Asymmetry

The use of digital technology in asset management involves an inherent symmetry paradox: it can enable more efficient and precise management while simultaneously amplifying existing asymmetries.

Technical approaches fall into two management models: “delegation type” and “autonomous type” [21]. This choice reflects an asymmetric trade-off between firms’ technological capabilities and their control rights. Delegated management can compensate for a firm’s limited technical capacity by relying on external specialists, but it may sacrifice data control and the security of core assets, producing an asymmetric transfer of rights. Conversely, autonomous management preserves control–right symmetry but demands substantial resources and high technical thresholds, thereby amplifying disparities in firm capabilities.

From a functional perspective, new technologies such as blockchain-based rights confirmation, artificial intelligence and machine learning, cloud-based content distribution systems, big data analytics, and digital identity infrastructures can theoretically establish a more symmetrical and transparent trust mechanism through means like traceability and rights confirmation [22]. Song Lvhuang’s (2016) exploration of the technical model for digital copyright protection is precisely an attempt to address the information and trust asymmetry between creators and users through technical means [23]. However, the other side of technology is that algorithmic bias may lead to an “information cocoon” in content distribution, exacerbating the asymmetry of information access. Lendvai and Gosztonyi (2025) demonstrate that algorithmic bias is a systemic outcome of human cognitive biases embedded in AI systems, which AI then amplifies, and that existing legal frameworks, including the GDPR, DSA, and AI Act, lack sufficient enforcement capacity to address the disproportionate impact on marginalized groups, while the “black box” nature of deep learning systems further obstructs the identification and legal challenge of discriminatory patterns [24]. If advanced digital protection technologies such as DRM are not used properly, they may evolve from tools for protecting rights into barriers that hinder universal access to knowledge, reinforcing the asymmetry in knowledge acquisition. Estrada’s (2010) research on the symmetry of complex networks further reveals that the digital ecosystem shaped by technological applications itself possesses inherent structural symmetry and breaking laws, which provides an analytical tool for understanding how technology systematically shapes or reverses the asymmetry of market structures [25]. The assessment of technology is therefore most informatively situated within the framework of how it affects the symmetry of the system.

2.4. Limitations of Research on Influencing Factors and Integration of “Symmetry” Framework

Previous studies examining factors that influence digital asset management have approached information asymmetry from a range of angles without integrating these findings systematically. Work framed around asset value connotation [20] or marketing [26] largely addresses micro-operational issues. Zhu (2020), for example, examined digitalization in the newspaper industry and identified timeliness of communication, content richness, and interactivity as key determinants of dissemination influence [27]. These elements primarily shape the quality of interactions between communicators and audiences, and thus affect the symmetry of information flow at the content level.

The existing literature has three interconnected limitations. First, its analytical perspective is fragmented: studies tend to isolate market structure, firm resources, or technological applications without integrating the asymmetries and intrinsic links among environmental, economic, and social dimensions into a single analytical framework. Second, a coherent theoretical lens is missing: prior work has not adopted “symmetry” as a core organizing concept to explain systematically the logical relationships among government regulation, digital asset management, and sustainable development outcomes. Third, and crucially, prior research has addressed information asymmetry predominantly from the supply side—examining how regulatory interventions expand information availability—while underweighting the demand side. Users differ significantly in their capacity and willingness to acquire, process, and update information, a phenomenon closely associated with the concept of Bounded Rationality [28]. Even when supply-side regulatory interventions successfully enhance information provision, demand-side heterogeneity in information literacy, updating behavior, and cognitive inertia may sustain residual asymmetries. A comprehensive understanding of whether and how regulatory efforts can move the system from asymmetry toward equilibrium therefore requires acknowledging both the supply-side and demand-side dimensions of information asymmetry. The present study focuses on supply-side regulatory mechanisms—an empirically tractable starting point—while explicitly positioning demand-side heterogeneity as a frontier for future research (see Section 6.3).

Recent advances in digital sustainability research make addressing these gaps feasible. Azizi E et al. (2025) examine digitalization as a driver of net-zero transition in local power systems [29], and Dobson A. (1996) synthesize evidence and pathways by which digital technologies promote resource optimization and economic sustainability [30]. Together, these studies imply that sustainable transformation is fundamentally a systematic process of correcting multiple asymmetries. In particular, Kshetri et al. (2022) review emerging technologies and resource challenges, directly examining environmental asymmetries associated with the energy consumption of technologies such as blockchain while also discussing rights and efficiency improvements; this work offers an important reference for building the integrated framework developed in this study [31].

In summary, the existing literature offers a strong foundation for understanding issues in digital asset management. Nonetheless, it lacks an integrated theoretical framework that systematically diagnoses multi-dimensional asymmetries and conceptualizes government regulation as a mechanism for restoring symmetry. Addressing this gap, the present study advances the core proposition “from asymmetry to symmetry.” This study empirically examines how three regulatory tools—resource allocation, technological innovation, and rights protection—interact synergistically in the digital asset management of the publishing and media industry to steer the system toward an efficient, equitable, and sustainable symmetrical state.

2.5. Symmetry Theory: Disciplinary Roots and an Operational Definition for Platform Governance

The concept of symmetry has been developed across multiple disciplines, each offering a distinct but complementary lens for understanding systemic balance in complex sociotechnical systems. Two theoretical traditions are directly relevant to platform governance, and their synthesis with contemporary platform studies provides the conceptual foundation for this study.

In complex systems science, symmetry refers to the invariance of a system’s structural properties under transformation. Garlaschelli et al. (2010) demonstrated that real-world complex networks exhibit inherent symmetry structures and characteristic patterns of symmetry-breaking, in which nodes occupy non-equivalent positions in terms of connectivity and influence [26]. Applied to digital ecosystems, this perspective reveals a critical diagnostic insight: when dominant platforms accumulate disproportionate connectivity, data flows, and market power, the network’s structural symmetry is broken. The resulting asymmetric topology is not a natural or neutral outcome but a structural condition that concentrates value creation and governance power among a small number of nodes—precisely the pattern documented in platform governance research. Gawer (2014) identifies this as a fundamental tension in digital platform ecosystems: external governance efforts face an inherent asymmetry with internal platform decision-making processes, as private platform companies become dominant power centers that disrupt the balance among network participants [32]. De Reuver et al. (2018) further demonstrate that digital platforms’ complex, distributed, and multi-sided nature creates layered governance challenges that cannot be addressed by any single regulatory instrument, underscoring the need for multi-dimensional frameworks capable of targeting distinct sources of structural imbalance [33].

In science and technology studies, Callon’s (1984) generalized symmetry principle holds that the analysis of sociotechnical networks should treat human actors, institutions, and nonhuman entities—technologies, standards, algorithms—on equal interpretive footing [14]. Applied to digital platform governance, this principle has a direct methodological implication: regulatory analysis should not privilege economic actors over technological artifacts, or vice versa. Rather, it should examine how resources, technological tools, and institutional rules together co-produce the asymmetric distributions of power that characterize contemporary platform ecosystems. This principle directly licenses the three-dimensional framework developed in Section 3: technology and rights protection are treated as independent governance dimensions—rather than secondary support functions—because algorithmic architectures and institutional rules generate distinct forms of asymmetry that require their own targeted corrective mechanisms.

These traditions are complementary rather than competing. The complex-systems perspective identifies structural asymmetry—where it exists, how it propagates through network topology, and why dominant platform actors are structurally positioned to resist redistributive pressures. The STS perspective identifies relational asymmetry—how the interactions among human actors, technologies, and institutions continuously reproduce imbalance. Together, they establish that diagnosing and correcting asymmetry in a platform ecosystem requires attending simultaneously to the structural properties of the system and to the governance mechanisms that shape its relational dynamics.

Building on this synthesis, this study defines platform governance symmetry as a systemic condition in which the structural distribution of resources, technological capabilities, and institutional rights across platform actors is sufficiently balanced to enable equitable participation, competitive fairness, and sustainable value creation for all stakeholders in the digital asset ecosystem. This definition is explicitly multi-dimensional—encompassing resource symmetry, capability symmetry, and rights symmetry—corresponding to the three regulatory dimensions developed in Section 3. It is also explicitly dynamic: symmetry here describes the directional property of the system’s trajectory rather than a fixed equilibrium point. The governance objective is therefore framed as achieving dynamic equilibrium—the operational state toward which symmetry-restoring regulatory interventions collectively direct the platform ecosystem. The relationship between the two concepts is sequential and directional: achieving symmetry across the three regulatory dimensions is the mechanism through which the system is steered toward dynamic equilibrium as its goal.

2.6. Sustainable Development in the Digital Economy: Conceptual Scope and This Study’s Position

Sustainable development is widely understood as a multi-dimensional concept encompassing three interrelated pillars: environmental sustainability, concerned with maintaining ecological quality and carrying capacity; social sustainability, encompassing human rights, equity, and the preservation of cultural diversity; and economic sustainability, concerned with the long-term maintenance of productive capital required for income and well-being [34].

The challenge is that these pillars generate trade-offs. Klarin (2018) explicitly warns that progress in one sustainability dimension may come at the cost of another [34]. Digital platform growth may advance economic efficiency while generating ecological externalities and deepening social inequalities simultaneously—producing what we term systemic asymmetry across sustainability dimensions.

Manioudis and Meramveliotakis (2022) further argue that dominant frameworks, including the SDGs, tend to reduce economic sustainability to GDP-oriented metrics, stripping the concept of the interdisciplinary depth that rigorous analysis requires [35]. They propose that sustainable development should be examined at two analytical levels: a transhistorical level identifying broad structural tendencies, and a historically specific level examining how those tendencies manifest within particular socioeconomic formations. This dual-level logic directly informs the present study, which examines how systemic asymmetries in China’s publishing and media industry generate unsustainable distributional outcomes and respond to targeted regulatory correction.

The present study’s empirical analysis focuses on the economic and social-institutional dimensions of digital sustainability—specifically, the efficiency and equity of content dissemination and the distribution of resource, technological, and rights-relational asymmetries. The environmental dimension is not directly measurable in our dataset, which contains no variables capturing energy consumption, carbon emissions, or green technology adoption. All empirical claims therefore pertain exclusively to the economic–institutional dimensions of sustainability. Environmental sustainability appears only as contextual background drawn from the broader literature [5], not as a finding derived from our analysis. We return to this scope boundary in the Discussion (Section 5) and Limitations (Section 6.3) Sections.

3. Theoretical Framework and Hypotheses

3.1. Theoretical Model

The definition of platform governance symmetry established in Section 2.5 provides the conceptual foundation for the analytical framework developed here. Two terms require precise disambiguation before the model is presented, as they operate at different levels of analysis throughout this study.

Symmetry, as used here, is a structural property of the platform ecosystem: it describes the degree to which resources, technological capabilities, and institutional rights are distributed in a balanced and equitable manner across actors in the digital asset value chain. Symmetry is the object of diagnosis: it can be assessed, decomposed into its three constituent dimensions, and used to identify which forms of imbalance are most acute and which regulatory interventions are best targeted to address them.

Dynamic equilibrium, as used in this study’s title and conclusions, is an outcome state: the condition a platform ecosystem approaches when governance mechanisms successfully and continuously correct the structural asymmetries that generate imbalance. “Dynamic” is a critical qualifier: unlike a static equilibrium in neoclassical economics—where forces are balanced and the system is at rest—dynamic equilibrium in a complex platform ecosystem denotes an ongoing process of regulatory adjustment in which the system remains responsive to new asymmetries as they emerge from technological change, market concentration, and evolving rights disputes. The relationship between the two concepts is therefore sequential and directional: achieving symmetry across the three regulatory dimensions is the mechanism through which the system is steered toward dynamic equilibrium as its goal. This distinction is reflected in the structure of this paper: the empirical analysis (Section 4 and Section 5) measures the symmetry-restoring effects of each regulatory dimension, while the governance strategy (Section 6) addresses the conditions under which their synergistic combination produces dynamic equilibrium at the system level. We acknowledge that dynamic equilibrium as defined here cannot be directly observed within the cross-sectional design of this study. What we can observe is the relative strength and complementarity of the three regulatory dimensions’ effects at a single point in time, which serves as an indirect indicator of whether the governance architecture is structurally capable of supporting an equilibrium-restoring process. Directly testing this process would require longitudinal or panel data—a task we explicitly recommend for future research (Section 6.3).

Amid a global reconfiguration of digital economic momentum, the digital transformation of China’s publishing and media industry has penetrated every link of the industrial chain. Yet this shift, while aimed at improving efficiency and growth, has not automatically produced a balanced or healthy industrial ecosystem; instead, it has exposed significant systemic asymmetries that extend beyond uneven technological adoption to the core structures of resource allocation, market power, and information flows, constituting the primary barrier to sustainable digital asset management.

3.1.1. Three Structural Asymmetries in China’s Publishing and Media Ecosystem

In the 5G and AI-driven digital operating environment, publishing and media companies are actively exploring new ways to disseminate knowledge and generate revenue. However, significant structural “blockages” remain across the industry chain—from content providers and publishers to service providers, platform operators, and end consumers. These blockages arise from the cumulative effect of three distinct asymmetries:

First, technological empowerment is asymmetric. Traditional publishing and media enterprises lag behind large Internet platforms in applying new technologies such as blockchain, smart contracts, and AI-driven distribution systems. This capability gap weakens their ability to trace, confirm rights to, and efficiently manage digital assets, preventing them from establishing independent, controllable management systems. Research on the digital divide demonstrates that such technology access disparities are structurally reproduced: McNeely (2024) shows that digital inequality stems from differential access to, use of, and benefit from ICT and AI across populations and regions, further compounded by social, cultural, economic, and political stratification factors that resist policy correction [36]. Gilbert (2010) similarly demonstrates that digital capability gaps are spatially embedded and multi-dimensional, requiring place-specific and capital-sensitive analytical frameworks to diagnose their drivers [37]. In the Chinese media context, this structural capability asymmetry directly constrains smaller and regional enterprises’ ability to compete with dominant platform operators.

Second, content adaptation and market power are asymmetric. Emerging content networks—exemplified by leading platforms such as WeChat, Weibo, and TikTok and MCN institutions—have changed the rules of asset operation. Traditional media enterprises retain a production logic that mismatches the industry’s need for rapid iteration and high interactivity. Consequently, media firms that specialize in professional fields confront a structural imbalance in market power when competing with comprehensive platforms that command vast traffic and algorithmic control, lacking bargaining strength in content distribution and revenue sharing. Cusumano MA et al. (2019) identify this as an inherent feature of multi-sided platform business models: the platform’s structural design creates market power asymmetries that conventional antitrust frameworks, based on static market definition, are ill-equipped to address [38]. Sun et al. (2017) further demonstrate that high market share among platform enterprises does not straightforwardly translate into competitive harm, because platform-mediated markets maintain dynamic competitive pressures through network effects, yet this very dynamism accelerates the structural divergence between platform leaders and followers [39].

Third, an asymmetry exists between information and rights. The prevalence of copyright infringement in digital environments stems from a stark imbalance in information flow: in complex online environments, original content often escapes the control of rights holders. High enforcement costs and the difficulty of gathering evidence create a persistent mismatch between rights and responsibilities, discouraging original creators and undermining the sustainable production of high-quality digital assets. This asymmetry is structurally entrenched: Uddin, M.H. et al. has documented the systematic challenges of rights enforcement in Chinese publishing enterprises [19], while Rahman (2023) demonstrates that digital copyright enforcement faces compounding obstacles—informational asymmetry, jurisdictional complexity, and technical limitations of DRM systems—that international treaty frameworks and industry licensing initiatives have yet to overcome [40]. The result is a governance gap in which rights holders, particularly smaller content producers, bear disproportionate enforcement costs relative to platform operators.

Currently, the industrial ecosystem exhibits an unbalanced and asymmetrical state across three dimensions: resources (resource allocation asymmetry), technology (capability empowerment asymmetry), and institutional rights (rights–responsibility asymmetry). To systematically diagnose and address these challenges, this study develops an analytical framework grounded in symmetry theory, structured around the core processes of asset management. As illustrated in Figure 1, the complete chain from content production to consumption encompasses these three core dimensions, which interact throughout the entire process and correspond to three key mechanisms for addressing the identified asymmetries: resource allocation seeks to mitigate resource endowment imbalances; technological innovation aims to bridge capability gaps; and rights protection strives to rectify the asymmetry between rights and obligations, risks and returns.

Based on a literature review, an analysis of the current state of digital assets, and comparative case studies, this study proposes a three-tier framework—resources, technology, and rights protection—and uses it to examine how six factors affect the dissemination of digital content: the full-network traceability system, the status of resource evidence collection, laws and regulations, audience awareness education, digital protection technologies, and the state of rights protection alliances. These relationships are summarized in Figure 2.

3.1.2. From Symmetry as Structural Property to Observable Enterprise Heterogeneity

Symmetry, in the platform-governance sense defined in Section 2.5, is a structural property of the distribution of resources, technological capabilities, and institutional rights across actors in the platform ecosystem. It is a distributional concept, not a per-firm quantity. We therefore do not claim that any single variable “measures symmetry” at the firm level. Instead, each variable in our analysis captures a specific, observable dimension of enterprise heterogeneity whose distribution across the sampled enterprises constitutes the empirical shadow of symmetry in that dimension.

Following this logic, we operationalize our three regulatory dimensions as follows. Resource symmetry is approached through content quantity (X3). The cross-enterprise distribution of X3 reflects the structural asymmetry of content production capacity: a more balanced distribution of X3 across enterprises corresponds to greater resource symmetry in the ecosystem, while a more concentrated distribution indicates structural resource asymmetry. Regulatory interventions—for example, public digital resource platforms—that narrow the X3 gap between top and bottom ranked enterprises are interpreted as symmetry-restoring in the resource dimension. Capability symmetry is approached through content satisfaction (X7), the technology-driven output quality measure. The cross-enterprise distribution of X7 reflects the structural asymmetry in technological capability translated into content quality output; narrowing this distribution corresponds to symmetry restoration in the technological dimension. Rights-relational symmetry is approached through engagement metrics, particularly likes count (X12). Because rights protection in digital environments materializes through interactional trust—healthy creator–platform–user relationships generate engagement—the cross-enterprise distribution of X12 indirectly reflects the health of the rights-relational architecture. We acknowledge that X12 is an imperfect proxy for rights symmetry, and we treat conclusions drawn from this proxy as appropriately conservative.

A clarification on analytical use of the term “symmetry” is warranted. Throughout this paper, “symmetry” is used in an analytical sense when referring to distributional properties diagnosable through cross-enterprise variance and correlation patterns. When the term appears in the policy discussion to describe the normative goal of regulatory reform, this represents an extension of the analytical diagnosis to policy vocabulary—not a second, looser meaning.

3.2. Measurement Framework and Research Hypotheses

The study compiled influence rankings for China’s publishing and media enterprises over the period 24 January 2025 to 7 December 2025. Using the Xinbang Index, the top 10 publishing and media enterprises were identified and ranked; the results are presented in

Table 1. Quantitative Ranking List 10 of New Content Influence of Publishing and Media Enterprises.

Ranking	Name	Subordinate	Xinbang Index A1	After the Xinbang Index Is Unitary A2 (A1/1000)	Comprehensive Influence A3
1	Southern Plus Client	Southern Media Group, Guangzhou, China	872.8	87.28%	1.8676
2	Live Nanyang Cloud Broadcast Station	Nanyang Media Group Co., Ltd., Nanyang, China	844.5	84.45%	1.8401
3	Nanyang Press Media	Nanyang Press Media, Nanyang, China	818.9	81.89%	1.8032
4	Huaihai Evening News	Huaihai Evening News Agency, Huai‘an, China	796.6	79.66%	1.7760
5	Nanyang Daily	Nanyang Daily Press, Nanyang, China	778.4	77.84%	1.7596
6	JiaShang Media	JiaShang Media MCN, Chengdu, China	769.6	76.96%	1.7549
7	Literacy Little Bookworm	National Academy of Governance Audiovisual Publishing House, Beijing, China	416	41.60%	1.2410
8	Snail’s Journey to the West	Leshan Snail Culture Communication Co., Ltd., Leshan, China	541.9	54.19%	0.9786
9	ZhiGeng Library	China Machine Press, Beijing, China	348.1	34.81%	0.6856
10	People’s Education Press Tik Tok Channel	People’s Education Press, Beijing, China	407	40.70%	0.6604

Note: Data sources for A1, A2, A3: Xinbang Index.

.

This paper examines how digital content dissemination affects asset management in publishing and media enterprises using the six evaluation dimensions described above. To quantify the relationship between each dimension and the influence ranking, we calculate Spearman’s rank correlation coefficient. Spearman’s rho measures the monotonic association between two ranked variables regardless of their distributions or sample size. The closer the absolute value of ρ (−1 $\le \mathsf{\rho }\le 1$) is to 1, the stronger the correlation (with −1 indicating a perfect negative correlation and 1 indicating a perfect positive correlation). In this study, we compute the correlations among the ranking sequences for the six factors using the following formula [41]:

$$\rho =1-{\displaystyle \frac{6{\displaystyle {\sum }_{i=1}^N}{(X_i-y_i)}^2}{N^3-N}}$$

Scholars such as Zhang Haitao have quantitatively analyzed content dissemination using five dimensions: content power, originality power, personality power, sharing power, and monetization power. Building on this framework, the study examines how six factors affect the breadth and depth of dissemination: the full-network traceability system, the status of resource evidence collection, laws and regulations, audience awareness education, digital protection technology, and rights protection alliances. This analysis clarifies each factor’s influence on digital asset management within the publishing and media industry. Referring to this quantification concept for final-level evaluation indices and considering the specific features of content dissemination in publishing and media enterprises, the study derives a set of quantitative indicators of communication influence tailored to the industry. The quantitative indicators of communication influence in the publishing and media industry are presented in Table 2.

Table 2. Quantitative Indicators of Communication Influence in the Publishing and Media Industry.

Evaluation Content	Evaluation Factor	Evaluation Secondary Indicator	Six-Factor Correlation Measure	Factor Attribution	Remarks	Variable Level	Whether to Apply in Regression?
Communication Breadth (B)	Contact Force (B1)	User Count (X1)	Prerequisite for audience awareness and education	Rights Protection (CR2)	Number of users of the publishing media enterprise	Platform-level (TikTok user base)	No (zero variance)
		Safeguard for audience awareness education (X2)	Rights Protection (CR2)	Average daily user usage time	Safeguard for audience awareness education	Platform-level	No (zero variance)
		Source for evidence collection (X3)	Resources (CR1)	Total volume of content as of the evaluation period end date	Source for evidence collection	Enterprise-level	Yes (resource)
	Cognitive Force (B2)	Original Content Rate (X4)	Embodiment of digital protection	Technology (CT)	Percentage of original content to full text within the evaluation period	Enterprise-level	Yes (technology)
		Followers Count (X5)	Foundation for	Technology (CT)	Number of fans/subscribers	Enterprise-level	Yes (technology)
Communication Depth (D)	Satisfaction (D1)	Technical Satisfaction (X6)	Reflects the effectiveness of digital protection technology	Technology (CT)	Rating scores for easy access across different terminal devices, security & reliability, and esthetic design	Platform-level (TikTok platform rating)	No (zero variance)
		Content Satisfaction (X7)	Embodiment of digital protection	Technology (CT)	Content update frequency	Enterprise-level	Yes (technology)
		Service Satisfaction (X8)	Effect of the rights protection alliance	Rights Protection (CR2)	Feedback response rate—Ad insertion rate	Platform-level	No (zero variance)
	Satisfaction (D1)	User Stickiness (X9)	Element of the full-network traceability system	Technology (CT)	Frequency of use within the evaluation period (default is number of works)	Enterprise-level	Yes (technology)
		Recommendation Ratio (X10)	Standard for laws and regulations usage	Rights Protection (CR2)	Registration rate from recommendations to other netizens	Platform-level	No (zero variance)
		Usage Intention (X11)	Standard for laws and regulations usage	Rights Protection (CR2)	Time from initial use to uninstall within the evaluation period	Platform-level	No (zero variance)
	Participation (D3) (Using Tik Tok as an example)	Likes Count (X12)	Promotion of audience awareness education	Rights Protection (CR2)	Number of likes within the evaluation period	Enterprise-level	Yes (rights)
		Comments Count (X13)	Promotion of audience awareness education	Rights Protection (CR2)	Number of comments within the evaluation period	Enterprise-level	Yes (rights)
		Shares/Reposts Count (X14)	Promotion of audience awareness education	Rights Protection (CR2)	Number of shares/reposts within the evaluation period	Enterprise-level	Yes (rights)

Note: Variables are classified into two levels. Platform-level context parameters reflect characteristics of the TikTok platform as a whole and are constant across all sampled enterprises by construction (e.g., aggregate user base, platform-wide satisfaction ratings). These are retained in Table 2 for contextual completeness but are excluded from regression analyses because they provide no differential explanatory power. Enterprise-level comparable metrics vary across the ten sampled enterprises and constitute the analytical basis for the regression and rank-correlation analyses.

Table 2. Quantitative Indicators of Communication Influence in the Publishing and Media Industry.

Evaluation Content	Evaluation Factor	Evaluation Secondary Indicator	Six-Factor Correlation Measure	Factor Attribution	Remarks	Variable Level	Whether to Apply in Regression?
Communication Breadth (B)	Contact Force (B1)	User Count (X1)	Prerequisite for audience awareness and education	Rights Protection (CR2)	Number of users of the publishing media enterprise	Platform-level (TikTok user base)	No (zero variance)
		Safeguard for audience awareness education (X2)	Rights Protection (CR2)	Average daily user usage time	Safeguard for audience awareness education	Platform-level	No (zero variance)
		Source for evidence collection (X3)	Resources (CR1)	Total volume of content as of the evaluation period end date	Source for evidence collection	Enterprise-level	Yes (resource)
	Cognitive Force (B2)	Original Content Rate (X4)	Embodiment of digital protection	Technology (CT)	Percentage of original content to full text within the evaluation period	Enterprise-level	Yes (technology)
		Followers Count (X5)	Foundation for	Technology (CT)	Number of fans/subscribers	Enterprise-level	Yes (technology)
Communication Depth (D)	Satisfaction (D1)	Technical Satisfaction (X6)	Reflects the effectiveness of digital protection technology	Technology (CT)	Rating scores for easy access across different terminal devices, security & reliability, and esthetic design	Platform-level (TikTok platform rating)	No (zero variance)
		Content Satisfaction (X7)	Embodiment of digital protection	Technology (CT)	Content update frequency	Enterprise-level	Yes (technology)
		Service Satisfaction (X8)	Effect of the rights protection alliance	Rights Protection (CR2)	Feedback response rate—Ad insertion rate	Platform-level	No (zero variance)
	Satisfaction (D1)	User Stickiness (X9)	Element of the full-network traceability system	Technology (CT)	Frequency of use within the evaluation period (default is number of works)	Enterprise-level	Yes (technology)
		Recommendation Ratio (X10)	Standard for laws and regulations usage	Rights Protection (CR2)	Registration rate from recommendations to other netizens	Platform-level	No (zero variance)
		Usage Intention (X11)	Standard for laws and regulations usage	Rights Protection (CR2)	Time from initial use to uninstall within the evaluation period	Platform-level	No (zero variance)
	Participation (D3) (Using Tik Tok as an example)	Likes Count (X12)	Promotion of audience awareness education	Rights Protection (CR2)	Number of likes within the evaluation period	Enterprise-level	Yes (rights)
		Comments Count (X13)	Promotion of audience awareness education	Rights Protection (CR2)	Number of comments within the evaluation period	Enterprise-level	Yes (rights)
		Shares/Reposts Count (X14)	Promotion of audience awareness education	Rights Protection (CR2)	Number of shares/reposts within the evaluation period	Enterprise-level	Yes (rights)

Note: Variables are classified into two levels. Platform-level context parameters reflect characteristics of the TikTok platform as a whole and are constant across all sampled enterprises by construction (e.g., aggregate user base, platform-wide satisfaction ratings). These are retained in Table 2 for contextual completeness but are excluded from regression analyses because they provide no differential explanatory power. Enterprise-level comparable metrics vary across the ten sampled enterprises and constitute the analytical basis for the regression and rank-correlation analyses.

Based on the symmetry framework developed in Section 3.1 and the operationalization structure in Section 3.1.1, we derive three research hypotheses. Unlike single-dimension positive-effect predictions, these hypotheses are designed to capture the specific cross-dimensional, conditional, and tier-heterogeneous predictions that symmetry theory—rather than general economic intuition—uniquely generates. Each hypothesis is directly testable by the empirical analyses in Section 4.2 and Section 4.3.

H1 (Hierarchical contribution):

The three regulatory dimensions contribute unequally to content dissemination efficiency. Specifically, within an integrated model, the standardized effect size of technological innovation (represented by content satisfaction, X7) is expected to exceed the standardized effect sizes of both resource allocation (X3) and rights protection (X12). This hierarchy—rather than the simple positive sign of any single dimension—constitutes the symmetry theoretic prediction: technology functions as the primary symmetry-restoring lever because capability asymmetry represents the most structurally binding constraint in a platform ecosystem characterized by rapid digital transformation, while resource allocation and rights protection operate in supporting rather than leading roles.

The rationale for this ordering is grounded in the three-tier asymmetry diagnosis in Section 3.1: technological empowerment asymmetry is identified as the most acute barrier because it determines whether an enterprise can translate resource inputs into dissemination outcomes at all. Without sufficient technological capability, additional resources yield diminishing returns and rights protection frameworks remain operationally underutilized. Content satisfaction (X7) is used as the representative technology indicator because it directly captures the output of technological capability at the content level—the interface at which platform asymmetry is most consequential for dissemination influence. H1 is tested via the standardized Beta coefficients of the integrated cross-dimensional regression (Section 4.2.4, Table 7).

H2 (Conditionality of rights protection):

The effect of rights protection on content dissemination efficiency is conditional rather than unconditional. Rights protection is expected to be statistically detectable as an independent driver only when complementary inputs from the technology and resource dimensions are sufficiently developed. Specifically, the coefficient of the rights protection indicator (X12) is expected to attenuate—or become statistically marginal—when technology (X7) and resource allocation (X3) are simultaneously controlled in an integrated model, relative to its effect in the dimension-specific model. This attenuation reflects the synergistic rather than autonomous character of rights-based institutional effects: a robust rights protection environment fosters the trust and participation incentives that translate into dissemination gains, but these gains are only realizable when the content quality and resource base are adequate to sustain meaningful creator and audience engagement.

Engagement indicators—primarily likes count (X12) and comments count (X13)—serve as observable proxies for the health of the rights-relational architecture within the digital asset value chain, capturing the degree to which creator–platform–user relationships produce active participation rather than passive consumption. The government’s role in rights protection—strengthening copyright law, promoting rights protection alliances, and reducing enforcement friction—is expected to restore asymmetry in the rights dimension by lowering the costs borne disproportionately by smaller and less powerful content producers. H2 is tested by comparing the coefficient and significance of X12 between Table 6 (dimension-specific, CR2’ model) and Table 7 (integrated model): a predicted attenuation from marginal significance to non-significance constitutes confirmation of conditionality rather than refutation.

H3 (Tier-specific dominant asymmetry):

The dominant source of asymmetry that constrains content dissemination efficiency varies systematically by enterprise tier within the industry. For lower-ranked enterprises that have not yet crossed a minimum content-production threshold, the resource dimension constitutes the binding constraint: content quantity (X3) is expected to correlate more strongly with dissemination influence for these enterprises than for higher-ranked enterprises. For higher-ranked enterprises that have surpassed the resource threshold, the technology dimension becomes the more operative constraint—dissemination influence depends increasingly on content quality and technological sophistication rather than production volume. This tier-differentiated constraint pattern reflects the sequential nature of symmetry restoration: resource parity is a prerequisite condition that must be established before technology-driven quality differentiation becomes the marginal source of competitive symmetry.

H3 is explored through the cross-enterprise Spearman rank correlation analysis (Section 4.3, Table 8): the prediction is that the rank correlation between X3 and influence (Y) strengthens as enterprise tier declines, while the rank correlation between X7 and Y remains relatively stable across tiers. Given the small-sample limitations discussed in Section 4.1 and Section 4.3, we treat the Spearman evidence as suggestive rather than confirmatory with respect to H3; the hypothesis is advanced primarily to motivate the tier-differentiated interpretation of the Spearman patterns and to establish a theoretically grounded basis for the tiered policy recommendations in Section 6.2.

4. Methodology and Data

4.1. Dataset

Adopting China’s publishing and media industry as a focal case, we empirically test this framework using secondary platform analytics data compiled from the Xinbang Index (www.newrank.cn), an industry-standard composite metric that quantitatively evaluates content dissemination efficacy and overall influence of new media accounts across platforms including TikTok (Douyin), WeChat Official Accounts, and Weibo. This approach is consistent with established practice in digital media research, where platform-generated behavioral metrics—including content engagement, follower dynamics, and interaction patterns—provide objective, high-frequency measures of dissemination performance that are not subject to the self-report biases inherent in questionnaire-based data collection.

This study analyzes the top 10 publishing and media enterprises ranked by the Xinbang Index over the period 24 January 2025 to 7 December 2025. The enterprises and their comprehensive influence rankings are presented in Table 1; the complete dataset of 10 enterprises was used in all regression and Spearman rank correlation analyses reported in Tables 4–7.

It should be noted that the Likes Count variable (X12) records net incremental values within the evaluation window rather than cumulative totals. Negative values observed for two enterprises reflect net decreases in cumulative likes during the measurement period—a phenomenon consistent with platform content moderation activities or account restructuring—and are retained in the analysis as valid observations of platform engagement dynamics.

We note the methodological characteristic of this analytical sample explicitly. The ten enterprises analyzed here do not constitute a random sample drawn from a larger population; they are the full census of the top-tier publishing and media enterprises on the Xinbang Index for the evaluation period. Inference in this study is therefore descriptive and exploratory rather than population-generalizing. Our goal is to identify directional patterns in how the three regulatory dimensions relate to content dissemination efficiency among the most prominent enterprises in the sector, which themselves account for a disproportionate share of industry influence. Conventional sample size concerns associated with inferential regression apply to samples drawn from populations; they apply in attenuated form to a population census, where every unit in the relevant population is observed. That said, we emphasize that the statistical estimates reported here should be interpreted as directional and exploratory, and we explicitly recommend replication with broader multi-sector and multi-country samples in future research (see Section 6.3).

4.2. Regression Results

Drawing on the raw indicator data presented in Table 3, regression analyses were conducted in which the comprehensive influence index (Y) served as the dependent variable and the platform analytics indicators representing each of the three regulatory dimensions—resources (CR1), technology (CT), and rights protection (CR2)—served as independent variables. Prior to regression analysis, all candidate variables were subjected to variance screening. Variables exhibiting near-zero variance across the analytical sample—specifically, X1 (user count), X6 (technical satisfaction), X8 (service satisfaction), X10 (recommendation ratio), and X11 (usage intention)—were excluded from regression models on the grounds that constant or near-constant variables contribute no differential explanatory power and artificially inflate model fit statistics. This screening procedure follows standard practice in regression analysis with secondary platform data, where certain aggregate platform-level metrics are uniform across all sampled accounts by construction.

Table 3. A Detailed List of Quantitative Indicators of Communication Influence of the Top Ten Publishing and Media Enterprises.

Enterprise Tik Tok Identifier		Southern Plus Client	Live Nanyang Cloud Broadcast Station	Nanyang Press Media	Huaihai Evening News	Nanyang Daily	JiaShang Media	Literacy Little Bookworm	Snail’s Journey to the West	ZhiGeng Library	People’s Education Press Tik Tok Channel
Y	Comprehensive Brand Influence A3	1.8676	1.8401	1.8032	1.7760	1.7596	1.7549	1.2410	0.9786	0.6856	0.6604
CR1	Content Quantity X3	704	659	493	202	485	238	291	466	117	80
CT	Original Content Rate X4	44.71%	100%	100%	66.67%	63.94%	100%	100%	100%	0	100%
	Followers Count X5	800,797	660,506	169,158	270,473	127,699	1,609,012	167,019	937,661	84,464	95,978
	Technical Satisfaction X6	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Content Satisfaction X7	85	228	159	51	208	6	6	12	1	1
	User Stickiness X9	85	228	159	51	208	6	6	12	1	1
CR2	User Count X1 (Unit: 100 million)	4	4	4	4	4	4	4	4	4	4
	User Usage Time X2 (Unit: 100 million hours)	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986
	Service Satisfaction X8	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Recommendation Ratio X10	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Usage Intention X11	16	16	16	16	16	16	16	16	16	16
	Likes Count X12	8,171,226	7,019,176	2,960,657	1,262,576	1,233,251	850,086	2128	44,700	−492	−837
	Comments Count X13	435,549	203,656	2048	5125	2053	7408	7	586	0	11
	Shares/Reposts Count X14	320,900	318,445	22,232	257,091	34,004	1274	53	88	1	9

According to Analysis One point, Tik Tok users’ total usage time rose to 459.86 million hours during 25–30 January 2025, marking its strongest performance in January. Additionally, as of 8 February, daily active users exceeded 400 million; this figure is used as a reference in this article. X7 and X9 share the same original indicator (content update frequency/number of publications) and play different conceptual roles across analytical dimensions, while remaining numerically identical at the raw level. In the regression analysis, only one of them is retained (X9 did not enter any regression, whereas X7 did—this has been properly addressed). The Likes Count variable (X12) records net incremental values within the evaluation window rather than cumulative totals. The negative values observed for ZhiGeng Library (X12 = −492) and People’s Education Press (X12 = −837) reflect net decreases in cumulative likes during the measurement period—a phenomenon consistent with platform content moderation activities (e.g., removal of content that had previously accumulated likes) or account restructuring events within the evaluation window. These observations are retained as valid records of platform engagement dynamics. Sensitivity analysis confirms that substituting zero for these two values leaves all regression coefficients and significance levels substantively unchanged (change in X12 coefficient: <5 × 10⁻¹¹; change in p-value: <0.001).

Table 3. A Detailed List of Quantitative Indicators of Communication Influence of the Top Ten Publishing and Media Enterprises.

Enterprise Tik Tok Identifier		Southern Plus Client	Live Nanyang Cloud Broadcast Station	Nanyang Press Media	Huaihai Evening News	Nanyang Daily	JiaShang Media	Literacy Little Bookworm	Snail’s Journey to the West	ZhiGeng Library	People’s Education Press Tik Tok Channel
Y	Comprehensive Brand Influence A3	1.8676	1.8401	1.8032	1.7760	1.7596	1.7549	1.2410	0.9786	0.6856	0.6604
CR1	Content Quantity X3	704	659	493	202	485	238	291	466	117	80
CT	Original Content Rate X4	44.71%	100%	100%	66.67%	63.94%	100%	100%	100%	0	100%
	Followers Count X5	800,797	660,506	169,158	270,473	127,699	1,609,012	167,019	937,661	84,464	95,978
	Technical Satisfaction X6	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Content Satisfaction X7	85	228	159	51	208	6	6	12	1	1
	User Stickiness X9	85	228	159	51	208	6	6	12	1	1
CR2	User Count X1 (Unit: 100 million)	4	4	4	4	4	4	4	4	4	4
	User Usage Time X2 (Unit: 100 million hours)	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986	4.5986
	Service Satisfaction X8	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Recommendation Ratio X10	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%
	Usage Intention X11	16	16	16	16	16	16	16	16	16	16
	Likes Count X12	8,171,226	7,019,176	2,960,657	1,262,576	1,233,251	850,086	2128	44,700	−492	−837
	Comments Count X13	435,549	203,656	2048	5125	2053	7408	7	586	0	11
	Shares/Reposts Count X14	320,900	318,445	22,232	257,091	34,004	1274	53	88	1	9

According to Analysis One point, Tik Tok users’ total usage time rose to 459.86 million hours during 25–30 January 2025, marking its strongest performance in January. Additionally, as of 8 February, daily active users exceeded 400 million; this figure is used as a reference in this article. X7 and X9 share the same original indicator (content update frequency/number of publications) and play different conceptual roles across analytical dimensions, while remaining numerically identical at the raw level. In the regression analysis, only one of them is retained (X9 did not enter any regression, whereas X7 did—this has been properly addressed). The Likes Count variable (X12) records net incremental values within the evaluation window rather than cumulative totals. The negative values observed for ZhiGeng Library (X12 = −492) and People’s Education Press (X12 = −837) reflect net decreases in cumulative likes during the measurement period—a phenomenon consistent with platform content moderation activities (e.g., removal of content that had previously accumulated likes) or account restructuring events within the evaluation window. These observations are retained as valid records of platform engagement dynamics. Sensitivity analysis confirms that substituting zero for these two values leaves all regression coefficients and significance levels substantively unchanged (change in X12 coefficient: <5 × 10⁻¹¹; change in p-value: <0.001).

All candidate variables were subjected to a pre-specified three-stage screening procedure before entering any regression model.

Stage 1—Variance screening. Variables exhibiting zero variance across the analytical sample were excluded because constant or near-constant variables contribute no differential explanatory power and artificially inflate model-fit statistics. Five variables were excluded on this basis: X1 (user count, platform-level constant = 4 hundred million for all enterprises), X6 (technical satisfaction, platform-wide TikTok rating = 100% for all enterprises), X8 (service satisfaction, platform-level = 100%), X10 (recommendation ratio, platform-level = 100%), and X11 (usage intention, platform-level = 16 for all enterprises). As documented in the revised Table 2, these five variables are classified as platform-level context parameters that do not vary across enterprise accounts by construction, and their exclusion reflects the hierarchical structure of the data rather than ad hoc judgment.

Stage 2—Multicollinearity testing (VIF). Variance Inflation Factors (VIF) were computed for all remaining variables within each dimensional model. Variables with VIF > 5 were removed sequentially. Results: (a) Resource model (X3 only): VIF = 1.000 by construction; (b) Technology model (X4, X5, X7): VIF(X4) = 1.12, VIF(X5) = 1.03, VIF(X7) = 1.03—all well below the threshold, and hence no exclusion on collinearity grounds (X4 was retained at the initial model stage but subsequently excluded due to near-zero t-statistic, as reported in Table 5); (c) Rights model (X12, X13, X14): VIF(X12) = 6.079, VIF(X13) = 6.08, VIF(X14) = 3.37—X12 and X13 both exceed the VIF = 5 threshold, confirming collinearity; X13 was excluded following the sequential-removal rule (highest VIF first), yielding a post-exclusion model (CR2′) in which VIF(X12) = 6.08 with X13 was also re-tested; given the small sample and theoretical primacy of X12 as the direct rights-engagement proxy, X12 was retained as the representative variable.

Stage 3—Heteroscedasticity testing (Breusch–Pagan). The Breusch–Pagan (BP) test was applied to each dimensional model and to the integrated cross-dimensional model (Section 4.2.4). The BP test statistic is computed as n × R² from an auxiliary regression of squared residuals on the original predictors. Results: Resource model (X3): BP = 2.050, χ²crit (df = 1, α = 0.05) = 3.841, p = 0.188—homoscedastic; Technology model (X5, X7): BP = 3.874, χ²crit (df = 2, α = 0.05) = 5.991, p = 0.180—homoscedastic; Rights model (X12, X13): BP = 2.596, χ²crit (df = 2, α = 0.05) = 5.991, p = 0.349—homoscedastic; Integrated model (X3, X7, X12): BP = 5.741, χ²crit (df = 3, α = 0.05) = 7.815—homoscedastic. No model exhibits statistically significant heteroscedasticity; ordinary least squares standard errors are therefore valid for all reported models.

The three-stage procedure is uniformly applied across all models and is fully pre-specified; variable exclusion decisions are rule-governed rather than data-driven.

4.2.1. The Synergy and Non-Determinacy of Resource Allocation (CR1)

Table 4 shows that the regression for the resource dimension centered on content volume (X3) has limited overall significance: the F-test p-value is 0.0447. This indicates that a single resource allocation cannot, by itself, dominates the drivers of communication impact at this stage and provides foundational support for the resource dimension of Hypothesis H1 while revealing its limits. Although the coefficient on content volume (X3) is positive and statistically significant (0.001457, p < 0.05), confirming that greater resource reserves are a necessary basis for expanding influence, the weak overall significance suggests that in the current industry context simply reducing initial resource endowment asymmetries is insufficient. Without concurrent improvements in technological transformation capacity and market penetration, the “equilibrium-restoring” effect of resource allocation is constrained. Therefore, resource allocation must be coordinated with other regulatory tools to achieve substantial effects.

Table 4. Results of resource-related regression analysis.

Variable		Coefficient	Standard Error	t-Statistic	p-Value
CR1	constant C	0.892369	0.261999	3.406004	0.0093
	Number of contents X3	0.001457	0.000613	2.377976	0.0447
$\overline{R^2}$	0.34089		F-statistic	0.044687

Table 4. Results of resource-related regression analysis.

Variable		Coefficient	Standard Error	t-Statistic	p-Value
CR1	constant C	0.892369	0.261999	3.406004	0.0093
	Number of contents X3	0.001457	0.000613	2.377976	0.0447
$\overline{R^2}$	0.34089		F-statistic	0.044687

4.2.2. Core Drivers and Key Levers of Technological Innovation (CT)

Table 5’s regression results identify technological innovation as the primary “equilibrium-restoring” mechanism. After addressing multicollinearity, technology-driven user content satisfaction (X₇) has a positive and highly significant effect (coefficient 0.004, p < 0.05), strongly supporting Hypothesis H1. This implies that government incentives for technological innovation can meaningfully raise content quality and user experience, thereby directly mitigating asymmetries in development capacity. Technology thus serves as the most effective “equilibrium-restoring lever”: it helps firms escape homogeneous content competition and, by boosting satisfaction, strengthens user retention and loyalty. This improvement in satisfaction is the key driver that shifts digital asset management from crude resource accumulation toward refined, high-quality development.

Table 5. Technology-related regression results.

Variant		Ratio	Standard Error	t-Statistic	Confidence Probability	Variant		Ratio	Standard Error	t-Statistic	Confidence Probability
CT	constant C	0.941401	0.324143	2.904276	0.0272	CT′	After excluding multicollinearity
	Content originality X4	−0.027986	0.39051	−0.071666	0.9452		constant C	0.923868	0.19695	4.690881	0.0022
	Number of concerns X5	4.33 × 10−7	2.65 × 10−7	1.63223	0.1538		Number of concern X5	4.27 × 10−7	2.34 × 10−7	1.828226	0.1102
	Content satisfaction X7	0.004011	0.001429	2.807691	0.0308		Content satisfaction X7	0.003995	0.001306	3.059673	0.0183
$\overline{R^2}$	0.424084		F-statistic	0.104336		$\overline{R^2}$	0.505935		F-statistic		0.035175

Table 5. Technology-related regression results.

Variant		Ratio	Standard Error	t-Statistic	Confidence Probability	Variant		Ratio	Standard Error	t-Statistic	Confidence Probability
CT	constant C	0.941401	0.324143	2.904276	0.0272	CT′	After excluding multicollinearity
	Content originality X4	−0.027986	0.39051	−0.071666	0.9452		constant C	0.923868	0.19695	4.690881	0.0022
	Number of concerns X5	4.33 × 10−7	2.65 × 10−7	1.63223	0.1538		Number of concern X5	4.27 × 10−7	2.34 × 10−7	1.828226	0.1102
	Content satisfaction X7	0.004011	0.001429	2.807691	0.0308		Content satisfaction X7	0.003995	0.001306	3.059673	0.0183
$\overline{R^2}$	0.424084		F-statistic	0.104336		$\overline{R^2}$	0.505935		F-statistic		0.035175

The non-significance of X5 (followers count) is consistent with the observed rank inversion in the raw data—JiaShang Media holds the highest follower count (1,609,012) despite ranking sixth overall—suggesting that raw audience scale is an insufficient proxy for dissemination efficiency in the absence of high-quality content output.

4.2.3. Foundational and Synergistic Role of Rights Protection (CR2)

Table 6’s regression results support Hypothesis H2. After controlling for confounding factors, the advocacy dimension is marginally positively associated with its primary indicator, the number of user likes (X12) (p = 0.0543). This finding indicates that rights protection functions as a fundamental, synergistic “equilibrium-restoring” mechanism rather than a decisive driver. A robust rights protection environment—manifested in interactive behaviors such as likes and comments—establishes essential trust and positive incentives that help balance the asymmetry of rights among creators, platforms, and users. Nevertheless, the effect of rights protection is conditional: it is fully realized only when content quality (technology) improves and the resource base expands. Thus, rights protection operates more as a safeguard within the regulatory toolkit than as the primary engine for creating a fair environment.

Table 6. Rights-related regression results.

Variant		Ratio	Standard Error	t-Statistic	Confidence Probability	Variant		Ratio	Standard Error	t-Statistic	Confidence Probability
CR2	constant C	1.059515	0.16459	6.437281	0.0007	CR2′	After excluding multicollinearity
	number of likes X12	3.64 × 10−7	1.38 × 10−7	2.637309	0.0387		constant C	1.123954	0.164083	6.84993	0.0002
	Number of comments X13	−4.81 × 10−6	2.45 × 10−6	−1.960303	0.0977		number of likes X12	2.41 × 10−7	1.04 × 10−7	2.308551	0.0543
	number of forwards X14	−2.20 × 10−7	1.71 × 10−7	−1.289624	0.2447		Number of comments X13	−3.14 × 10−6	2.18 × 10−6	−1.440086	0.193
$\overline{R^2}$	0.454062		F-statistic	0.089823		$\overline{R^2}$	0.402344		F-statistic	0.068482

Table 6. Rights-related regression results.

Variant		Ratio	Standard Error	t-Statistic	Confidence Probability	Variant		Ratio	Standard Error	t-Statistic	Confidence Probability
CR2	constant C	1.059515	0.16459	6.437281	0.0007	CR2′	After excluding multicollinearity
	number of likes X12	3.64 × 10−7	1.38 × 10−7	2.637309	0.0387		constant C	1.123954	0.164083	6.84993	0.0002
	Number of comments X13	−4.81 × 10−6	2.45 × 10−6	−1.960303	0.0977		number of likes X12	2.41 × 10−7	1.04 × 10−7	2.308551	0.0543
	number of forwards X14	−2.20 × 10−7	1.71 × 10−7	−1.289624	0.2447		Number of comments X13	−3.14 × 10−6	2.18 × 10−6	−1.440086	0.193
$\overline{R^2}$	0.454062		F-statistic	0.089823		$\overline{R^2}$	0.402344		F-statistic	0.068482

The marginal significance of the rights protection dimension (p = 0.0685, α = 0.10) is itself theoretically informative. Rather than undermining the framework, this finding is consistent with H2’s prediction that rights protection operates as a synergistic institutional guarantee rather than an independent driver. The conditional nature of its effect—dependent on complementary technological and resource inputs—reflects the well-documented challenges of rights enforcement in digital environments [17]: legal and institutional mechanisms produce their full symmetry-restoring effects only when the underlying content ecosystem is sufficiently developed through technology and resource investment to make rights protection operationally meaningful.

In summary, the regression results indicate that guiding the publishing and media industry’s digital ecosystem from “asymmetry” to “symmetry” requires a coordinated approach. Technological innovation is the most effective lever for achieving dynamic equilibrium. Resource allocation provides the necessary synergistic foundation, and rights protection serves as the indispensable institutional guarantee. No single regulatory instrument can independently accomplish systemic “equilibrium-restoring”; instead, an integrated, synergistic policy framework is urgently needed.

4.2.4. Integrated Cross-Dimensional Regression: Robustness Check and Hierarchical Test of H1

The dimension-by-dimension regressions in Section 4.2.1, Section 4.2.2 and Section 4.2.3 were designed to identify the dominant variable within each regulatory dimension and to provide within-dimension diagnostics (collinearity, heteroscedasticity). However, they do not permit assessment of each dimension’s relative contribution while simultaneously controlling for the others. To address this limitation directly and to provide a formal test of H1 (hierarchical contribution), we estimate a single integrated regression model incorporating one representative variable from each dimension: X3 (content quantity, resource), X7 (content satisfaction, technology), and X12 (likes count, rights protection).

Table 7 reports the integrated model results. All VIF values are below the threshold of 5 (VIFX3 = 3.878, VIFX7 = 2.004, VIFX12 = 3.003), confirming the absence of serious multicollinearity among the three representative variables. The Breusch–Pagan test yields BP = 5.741, below the critical value of 7.815 (χ², df = 3, α = 0.05), confirming homoscedasticity. The integrated model achieves R² = 0.525 and Adjusted R² = 0.287. The overall F-test is F(3, 6) = 2.208, p = 0.188, which does not reach conventional significance. This outcome is fully anticipated given the census size of n = 10 with three predictors, which allows for only six residual degrees of freedom—a known constraint of inference in population censuses (see Section 4.1). Individual coefficient p-values are similarly wide (X3: p = 0.829; X7: p = 0.361; X12: p = 0.569), reflecting the limited power available rather than the absence of directional effects.The integrated cross-dimensional regression results are presented in Table 7.

Table 7. Integrated Cross-Dimensional Regression Results (Y~X3 + X7 + X12).

Panel A: Model Summary
R	R2	Adjusted R2	Std. Error of Estimate		Durbin–Watson
0.724	0.525	0.287	0.419		1.112
Panel B: ANOVA
	Sum of Squares	df	Mean Square		F	Sig.
Regression	1.164	3	0.388		2.208	0.188
Residual	1.054	6	0.176
Total	2.218	9
Panel C: Coefficients and Collinearity Statistics
Variable	B	Std. Error	β	t	Sig.	VIF
	(Unstd.)		(Std.)
Constant (β0)	1.0628	0.3246	—	3.274	0.017 **	—
X3 (Content Quantity)	2.833 × 10−4	1.255 × 10−3	0.125	0.226	0.829 n.s.	3.878
X7 (Content Satisfaction)	2.169 × 10−3	2.192 × 10−3	0.394	0.989	0.361 n.s.	2.004
X12 (Likes Count)	4.823 × 10−8	8.011 × 10−8	0.294	0.602	0.569 n.s.	3.003

Notes: Dependent variable: Comprehensive Brand Influence (Y). Independent variables are the representative indicators of the three regulatory dimensions: X3 (resource), X7 (technology), X12 (rights protection). All VIF values < 5, indicating no serious multicollinearity. Breusch–Pagan heteroscedasticity test: BP statistic = 5.741 < χ²crit (df = 3, α = 0.05) = 7.815, indicating homoscedasticity. n.s. = not significant; ** p < 0.05. n = 10. Standardized Beta (β) hierarchy: X7 (β = 0.394) > X12 (β = 0.294) > X3 (β = 0.125). This ordering is consistent with H1’s prediction that technological innovation constitutes the dominant symmetry-restoring lever, followed by rights protection and resource allocation. Although individual coefficients do not reach conventional significance thresholds—an expected consequence of the six degrees of freedom available in this n = 10 census—the directional hierarchy is unambiguous and theoretically coherent.

Table 7. Integrated Cross-Dimensional Regression Results (Y~X3 + X7 + X12).

Panel A: Model Summary
R	R2	Adjusted R2	Std. Error of Estimate		Durbin–Watson
0.724	0.525	0.287	0.419		1.112
Panel B: ANOVA
	Sum of Squares	df	Mean Square		F	Sig.
Regression	1.164	3	0.388		2.208	0.188
Residual	1.054	6	0.176
Total	2.218	9
Panel C: Coefficients and Collinearity Statistics
Variable	B	Std. Error	β	t	Sig.	VIF
	(Unstd.)		(Std.)
Constant (β0)	1.0628	0.3246	—	3.274	0.017 **	—
X3 (Content Quantity)	2.833 × 10−4	1.255 × 10−3	0.125	0.226	0.829 n.s.	3.878
X7 (Content Satisfaction)	2.169 × 10−3	2.192 × 10−3	0.394	0.989	0.361 n.s.	2.004
X12 (Likes Count)	4.823 × 10−8	8.011 × 10−8	0.294	0.602	0.569 n.s.	3.003

Notes: Dependent variable: Comprehensive Brand Influence (Y). Independent variables are the representative indicators of the three regulatory dimensions: X3 (resource), X7 (technology), X12 (rights protection). All VIF values < 5, indicating no serious multicollinearity. Breusch–Pagan heteroscedasticity test: BP statistic = 5.741 < χ²crit (df = 3, α = 0.05) = 7.815, indicating homoscedasticity. n.s. = not significant; ** p < 0.05. n = 10. Standardized Beta (β) hierarchy: X7 (β = 0.394) > X12 (β = 0.294) > X3 (β = 0.125). This ordering is consistent with H1’s prediction that technological innovation constitutes the dominant symmetry-restoring lever, followed by rights protection and resource allocation. Although individual coefficients do not reach conventional significance thresholds—an expected consequence of the six degrees of freedom available in this n = 10 census—the directional hierarchy is unambiguous and theoretically coherent.

Hierarchical test of H1: Standardized Beta analysis. While individual significance tests are uninformative under n = 10, the standardized Beta coefficients provide a scale-free comparison of relative contribution that is directly interpretable regardless of statistical power. The Beta hierarchy is unambiguous: βX7 = 0.394 > βX12 = 0.294 > βX3 = 0.125. This ordering—Technology > Rights Protection > Resource Allocation—is precisely the hierarchy H1 predicts on the basis of symmetry theory: technological capability asymmetry is the most binding constraint on dissemination efficiency, followed by rights-relational asymmetry, with resource asymmetry playing a foundational but less decisive role. The integrated model therefore provides directional confirmation of H1 at the level of relative magnitude, even though the sample size precludes conventional significance tests of individual coefficients.

Relationship to H2 (conditionality of rights protection). The attenuation of X12’s coefficient in the integrated model relative to the dimension-specific model (Table 6: B = 2.409 × 10⁻⁷, p = 0.054; integrated: B = 4.823 × 10⁻⁸, p = 0.569) is consistent with—and indeed predicted by—H2. H2 states that rights protection’s effect on dissemination efficiency is conditional on complementary technology and resource inputs; when technology (X7) and resource allocation (X3) are simultaneously controlled, the independent marginal effect of X12 diminishes, reflecting its synergistic rather than autonomous mechanism. The sign of X12’s coefficient remains positive in the integrated specification, confirming the directional prediction; only its independent marginal contribution, already shared with X7 and X3, is absorbed by the larger model.

Interpretation and limitations. The integrated model should be read as a robustness and hierarchical check that complements—not replaces—the dimension-specific analyses: the dimension-specific models reveal within-dimension variable selection and provide the primary inferential evidence for H1–H3 in their original formulation, while the integrated model tests whether the cross-dimensional ordering survives simultaneous control. Their convergence in directional pattern (Technology > Rights > Resource, consistently positive coefficients) strengthens our substantive conclusions despite the power constraints of the analytical sample.

4.3. Validation Analysis

To complement the regression-based analyses in Section 4.2.1, Section 4.2.2, Section 4.2.3 and Section 4.2.4, we report Spearman rank correlation coefficients as a nonparametric robustness check. We distinguish two analytical components: a descriptive per-enterprise pattern analysis (Table 8) and a formal cross-enterprise analysis (Table 9).

Table 8. Results of Spearman’s Rank Correlation Coefficient.

Enterprises Projects	Resource	Technology		Rights Protection
	Number of Contents X3	Number of Concerns X5	Content Satisfaction X7	Number of Likes X12	Number of Comments X13
Southern Plus Client	0.100	0.715	0.922	−0.797	0.986
Live Nanyang Cloud Broadcast Station	0.198	0.754	0.927	−0.750	0.997
Nanyang Press Media	0.460	0.780	0.958	0.236	1.000
Huaihai Evening News	0.606	0.782	0.973	0.767	1.000
Nanyang Daily	0.631	0.786	0.974	0.864	1.000
JiaShang Media	0.772	0.787	1.000	0.956	1.000
Literacy Little Bookworm	0.806	0.944	1.000	1.000	1.000
Snail’s Journey to the West	0.857	0.946	1.000	1.000	1.000
ZhiGeng Library	0.988	0.999	1.000	1.000	1.000
People’s Education Press Tik Tok Channel	0.996	0.999	1.000	1.000	1.000

Table 8. Results of Spearman’s Rank Correlation Coefficient.

Enterprises Projects	Resource	Technology		Rights Protection
	Number of Contents X3	Number of Concerns X5	Content Satisfaction X7	Number of Likes X12	Number of Comments X13
Southern Plus Client	0.100	0.715	0.922	−0.797	0.986
Live Nanyang Cloud Broadcast Station	0.198	0.754	0.927	−0.750	0.997
Nanyang Press Media	0.460	0.780	0.958	0.236	1.000
Huaihai Evening News	0.606	0.782	0.973	0.767	1.000
Nanyang Daily	0.631	0.786	0.974	0.864	1.000
JiaShang Media	0.772	0.787	1.000	0.956	1.000
Literacy Little Bookworm	0.806	0.944	1.000	1.000	1.000
Snail’s Journey to the West	0.857	0.946	1.000	1.000	1.000
ZhiGeng Library	0.988	0.999	1.000	1.000	1.000
People’s Education Press Tik Tok Channel	0.996	0.999	1.000	1.000	1.000

Formal cross-enterprise Spearman analysis (Table 9). Table 9 reports the standard Spearman rank correlations between each regulatory indicator and the composite influence score (Y) across all ten enterprises. Four of the five indicators exhibit significant positive associations with Y: X3 (ρ = 0.830, p = 0.003), X7 (ρ = 0.835, p = 0.003), X12 (ρ = 0.988, p < 0.001), and X13 (ρ = 0.830, p = 0.003). The single exception, X5 (followers count, ρ = 0.455, p = 0.187), is consistent with the regression finding that raw audience scale is not a reliable predictor of dissemination efficiency independent of content quality. These associations are directionally consistent with H1–H3: all significant indicators relate positively to influence, and the rights protection indicator X12 shows the strongest cross-enterprise rank association. We note, however, that rank correlations in a sample of n = 10 are sensitive to distributional extremes; the particularly high ρ for X12 reflects in part the skewed distribution of likes (the top two enterprises account for the large majority of total likes), and should be interpreted as suggestive rather than precise.

Table 9. Cross-Enterprise Spearman Rank Correlations between Regulatory Indicators and Composite Influence (Y).

Variable	Regulatory Dimension	Spearman ρ	p-Value	Significance
X3 (Content Quantity)	Resource	0.83	0.003	**
X5 (Followers Count)	Technology	0.455	0.187	n.s.
X7 (Content Satisfaction)	Technology	0.835	0.003	**
X12 (Likes Count)	Rights Protection	0.988	0	***
X13 (Comments Count)	Rights Protection	0.83	0.003	**

Note: Statistical significance levels are indicated as follows: ** p < 0.01; *** p < 0.001; n.s. indicates non-significance.

Table 9. Cross-Enterprise Spearman Rank Correlations between Regulatory Indicators and Composite Influence (Y).

Variable	Regulatory Dimension	Spearman ρ	p-Value	Significance
X3 (Content Quantity)	Resource	0.83	0.003	**
X5 (Followers Count)	Technology	0.455	0.187	n.s.
X7 (Content Satisfaction)	Technology	0.835	0.003	**
X12 (Likes Count)	Rights Protection	0.988	0	***
X13 (Comments Count)	Rights Protection	0.83	0.003	**

Note: Statistical significance levels are indicated as follows: ** p < 0.01; *** p < 0.001; n.s. indicates non-significance.

Descriptive per-enterprise pattern analysis (Table 8). Table 8 presents per-enterprise descriptive coefficients that characterize the degree of alignment between each enterprise’s own indicator profile and its overall influence rank. These coefficients are not standard cross-enterprise Spearman correlations; rather, they serve a descriptive function, illustrating how the strength of association between regulatory indicators and influence varies across competitive tiers. We report Table 8 descriptively and do not rely on it for inferential claims.

A lower-triangular pattern is visible in Table 8: coefficients are generally lower for higher-ranked enterprises and approach 1.000 for lower-ranked enterprises. We explicitly acknowledge that this pattern is in part a mathematical property of small-sample rank comparisons—with n = 10, extreme-ranked observations have limited room for discordance, producing coefficients near ±1 by construction. The pattern is therefore consistent with, but not conclusive evidence of, the tier-specific asymmetry dynamics predicted by H3. We treat it as illustrative context that motivates the H3 prediction rather than as independent confirmation.

For Southern Plus Client (ranked first), the per-enterprise coefficient for X12 is −0.797. We do not interpret this as a substantive finding. Given that Southern Plus Client occupies the top rank with the highest absolute X12 value (8,171,226 likes), the negative coefficient reflects a rank-comparison artifact at the distributional extreme rather than a meaningful inverse relationship. We acknowledge this as a limitation of the per-enterprise descriptive analysis rather than a property of the underlying data.

The cross-enterprise Spearman analysis (Table 9) provides nonparametric corroboration of the regression results: the positive and significant associations for X3, X7, X12, and X13 are directionally consistent with H1–H3. Primary inferential weight rests on the regression analyses (Table 4, Table 5, Table 6 and Table 7); the Spearman analysis serves as a directional robustness check.

5. Discussion

5.1. Discussion of Findings

The empirical results of this study yield both confirmations and advancements relative to the existing body of literature on digital governance and sustainable transformation.

5.1.1. Technological Innovation as the Pivotal Lever

While prior research, such as that by Song Lvhuang (2016), has established the role of digital technology in mitigating information and trust asymmetry [23], this study identifies technological innovation as the most potent “symmetrical” driver. The significant positive effect of technology-driven user content satisfaction (X7) on dissemination influence, robustly confirmed by both regression and Spearman’s rank correlation analyses, underscores that government promotion of content quality-enhancing technologies serves as the most effective mechanism to rectify “technological capability asymmetry.” This finding resonates with but critically extends the work of Batista J V. (2021), by not only highlighting technology’s role in operational efficiency but also empirically linking it to the core of symmetrical restructuring within the media industry’s asset management [42].

5.1.2. The Synergistic and Non-Deterministic Role of Resource Allocation

The findings regarding resource allocation reveal a more nuanced picture than traditionally assumed. Although content quantity (X3) shows a positive correlation with influence, the limited explanatory power of the resource-only regression model indicates that resource input alone is insufficient to drive systematic symmetry. This result partially supports but also qualifies the emphasis on resource endowments found in earlier studies [21]. It demonstrates that resource allocation primarily furnishes the essential initial conditions, particularly for small and medium-sized enterprises, and needs to be synergized with technological innovation to realize its full potential. From an environmental perspective, this underscores the importance of policies aimed at equitable access to digital infrastructure, which can prevent redundant, inefficient resource deployment and mitigate competitive asymmetries across enterprise tiers.

5.1.3. The Foundational Guarantee of Rights Protection

The role of rights protection as an institutional guarantee is clearly delineated in our results. While interaction metrics like the number of likes (X12) and comments (X13) showed positive correlations, their marginal significance in regressions suggests that rights protection operates by fostering a trustworthy environment and encouraging sustained participation, rather than directly boosting dissemination effectiveness. This finding aligns with and substantiates the theoretical arguments of scholars like Wu Jing (2015) regarding the high costs of rights enforcement in the digital environment [17]. It confirms that rights protection is a fundamental, synergistic mechanism for resolving the “asymmetry between rights and responsibilities,” creating the necessary stability for long-term investments—including those in sustainable practices—by content creators and distributors.

5.2. Theoretical Contributions

This study advances three interconnected theoretical contributions to the literature on digital platform governance and sustainable development.

First, this study is the first to integrate symmetry theory with a multi-dimensional sustainable development framework and apply this combined lens specifically to the governance of digital assets on media platforms. Prior scholarship has examined regulatory design, digital transformation, and sustainability largely in isolation; by synthesizing these streams through the construct of “systematic symmetry”, this study provides a novel and generative theoretical vocabulary for diagnosing the multi-dimensional asymmetries—economic, social–institutional, and, at the conceptual level, environmental—that characterize platform-dependent industries in the digital economy.

Second, this study moves beyond the fragmented, single-factor analyses that have dominated existing research by constructing and empirically validating an integrated analytical framework that reconceptualizes government regulation as a synergistic, multi-dimensional platform governance mechanism. Rather than treating resource allocation, technological innovation, and rights protection as independent policy levers, this framework reveals how they function as mutually reinforcing instruments within a coherent governance architecture. In doing so, it effectively bridges the logical relationships among regulatory design, firm-level digital asset management, and overarching sustainable development goals—relationships that prior work has addressed only in part.

Third, the empirical validation of the technology-centered, resource-supported, rights-guaranteed collaborative mechanism offers a refined and nuanced understanding of how regulatory tools interact to produce systemic equilibrium. The finding that technological innovation functions as the decisive “symmetrical” lever, while resource allocation and rights protection operate as necessary but insufficient complements, demonstrates that dynamic equilibrium on media platforms is not achievable through isolated intervention but requires strategic, synergistic policy combination. This insight extends symmetry theory from its origins in physics and complex systems into the domain of digital platform governance, suggesting it as a productive analytical framework for examining regulatory strategies in other platform-dependent sectors—a proposition that the present exploratory study motivates but cannot itself confirm, given its census-based, sector-specific design.

5.3. Practical Implications

The empirical findings of this study generate actionable implications for the principal stakeholders involved in governing and operating digital media platform ecosystems.

For regulatory authorities and policymakers, the results indicate that effective governance of digital asset platforms is more likely to succeed when it moves beyond isolated, single-instrument interventions toward an integrated, technology-centered approach. Concretely, this entails three coordinated lines of action: prioritizing technological innovation by directing R&D subsidies and procurement incentives toward content quality-enhancing technologies, inclusive algorithms, and intelligent content distribution infrastructure; optimizing resource allocation by establishing public digital resource platforms and ensuring equitable infrastructure access for underserved regions and small-to-medium enterprises; and fortifying the institutional rights protection framework by strengthening copyright enforcement mechanisms, facilitating cross-platform rights protection alliances, and embedding social equity criteria into regulatory impact assessments. The empirical evidence suggests that designing and implementing these three lines of action in concert is associated with greater symmetry-restoring effectiveness than any single dimension alone.

For publishing and media enterprises operating within platform ecosystems, the findings highlight the strategic imperative of proactive technological capacity-building as the primary means of mitigating competitive asymmetries against dominant platform actors. The evidence suggests that firms may benefit from prioritizing investment in advanced digital solutions for content protection, intelligent distribution, and platform-native user experience optimization. Such technological upgrading not only enhances content quality and audience engagement—directly strengthening market influence and operational sustainability—but also positions enterprises to benefit more fully from the resource allocation and rights protection support that regulatory frameworks provide.

For the pursuit of the United Nations Sustainable Development Goals (SDGs), this study offers relevant but bounded implications. The empirical findings directly support the economic and social–institutional dimensions of the SDG agenda: SDG 9 (Industry, Innovation and Infrastructure) is addressed by the finding that government-supported technological innovation is the primary lever for narrowing content quality capability gaps; SDG 10 (Reduced Inequalities) is addressed by the resource threshold pattern, which identifies smaller and lower-ranked enterprises as the group most constrained by resource asymmetry and therefore most in need of targeted support; and SDG 16 (Peace, Justice and Strong Institutions) is addressed by the rights protection findings, which demonstrate that institutional trust frameworks constitute an important enabling condition for sustainable platform participation. The environmental dimensions of the SDG agenda—including SDG 7 (Clean Energy) and SDG 13 (Climate Action)—while intrinsically linked to the long-term trajectory of digital infrastructure development, fall outside the empirical scope of the present study, as noted in Section 2.6. Policymakers are encouraged to align platform governance with these economic and social–institutional SDG dimensions that are directly addressable by the content dissemination mechanisms examined in this study and to treat the environmental dimension as a priority for future integrated assessment frameworks.

6. Conclusions

6.1. Research Conclusion

This study examined how government regulation shapes digital asset management in the publishing and media industry. It applied symmetry theory to assess whether three regulatory dimensions—resource allocation, technological innovation, and rights protection—can collectively move the industry from multi-dimensional asymmetry toward dynamic equilibrium. The findings confirm that each dimension contributes distinctly, but with notably different levels of potency.

Technological innovation is the most effective symmetry-restoring mechanism. It operates by improving content quality and user experience, directly narrowing capability gaps between smaller and dominant platform participants. Resource allocation is structurally foundational, but its independent effect on dissemination efficiency is limited. Its contribution is realized primarily in combination with technological inputs. Resource provision without technological enablement does not produce systemic equilibrium. Rights protection functions as a conditional institutional safeguard. It supports equitable participation by fostering trust among creators, platforms, and users. Its effect, however, is indirect and depends on adequate technology and resource conditions being in place.

Collectively, these findings indicate that no single regulatory instrument is sufficient to move the system toward dynamic equilibrium. A coordinated combination—where resource allocation provides the material foundation, technology drives capability convergence, and rights protection provides the institutional guarantee—is more likely to produce sustainable outcomes than any single dimension alone. Beyond the publishing and media sector, the symmetry framework provides an analytical template for examining analogous asymmetries in other platform-dependent industries. Whether the directional hierarchy observed here generalizes to other sectors and countries remains a productive hypothesis for future empirical work.

6.2. Policy Recommendations

The empirical results support a tripartite governance model. It aligns technological advancement, resource equity, and institutional integrity to support the media industry’s sustainable platform development.

First, establish technological innovation as the primary governance lever. Content satisfaction (X7) is the most robust predictor of dissemination influence across all three regulatory dimensions. This effect persists in both the dimension-specific regression (Table 5) and the integrated regression (Table 7). The evidence supports policies that target content quality disparities directly. Regulatory authorities may consider (a) technology adoption subsidies for small- and mid-tier enterprises, conditioned on measurable improvements in content quality metrics comparable to X7; (b) shared content production infrastructure—such as editing tools and AI-assisted services—to narrow the capability gap without requiring each firm to invest at full scale; (c) content quality monitoring frameworks to track whether interventions are, in fact, narrowing the cross-enterprise distribution of content satisfaction (Section 3.1).

Second, restructure resource allocation with tier-differentiated targeting. Content quantity (X3) is significantly associated with dissemination influence (Table 4). Its rank correlation with influence is stronger for lower-ranked enterprises (Table 8, with caveats noted in Section 4.3). This pattern is consistent with H3’s prediction of a resource threshold effect. Accordingly, the evidence supports differentiating policy by enterprise tier: (a) for smaller and lower-ranked enterprises, targeted support for baseline content-production capacity—through public platforms, shared infrastructure, and production grants—addresses the binding scarcity constraint; (b) for mid- and upper-tier enterprises, resource policy may shift from quantity-focused to quality-focused instruments, complementing the technology-centered priorities in Recommendation 1; (c) monitoring the cross-enterprise distribution of X3 provides an indicator of whether resource policy is narrowing distributional asymmetry rather than simply raising aggregate output.

Third, strengthen rights protection as a synergistic institutional foundation. The rights protection dimension contributes positively but at marginal significance in the dimension-specific regression (X12, p = 0.0543, Table 6). Its effect is more strongly shaped by interaction with the other dimensions than by independent operation (Table 7). This is consistent with H2’s conditionality prediction. Three evidence-supported implications follow: (a) bundling rights protection policies with technology and resource interventions is associated with greater effectiveness than isolated enforcement; (b) institutional infrastructure—copyright registries, rights protection alliances, and streamlined enforcement channels—is most effective when designed for low friction costs, enabling rights protection to activate once complementary technology and resource conditions are in place; (c) addressing the asymmetric enforcement costs faced by smaller creators and enterprises is a structural priority, as this asymmetry is a primary channel through which rights imbalance propagates [18].

Together, the three recommendations form a synergistic governance model. The empirical evidence suggests their effectiveness is greater when implemented in coordination than when applied individually.

Table 10. Empirical Source Map: Policy Recommendations and Their Evidential Basis.

Policy Sub-Recommendation	Corresponding Empirical Finding	Data Table
Rec. 1(a)(b): Technology-adoption support for small- and mid-tier enterprises	X7 (content satisfaction) is the most robust positive predictor in both dimension-specific and integrated regressions	Table 5 and Table 7
Rec. 1(c): Content quality distribution monitoring	Cross-enterprise variance in X7 is the direct operational indicator of technological capability asymmetry	Table 3 (X7 raw distribution)
Rec. 2(a): Resource targeting toward lower-ranked enterprises	Rank correlation between X3 and influence strengthens as enterprise tier declines, consistent with a resource threshold effect (interpreted with Section 4.3 caveats)	Table 4 and Table 8
Rec. 2(b): Quality-focused instruments for upper-tier enterprises	Marginal returns to X3 diminish for higher-ranked enterprises; X7 becomes the dominant predictor	Table 5 and Table 7 vs. Table 4
Rec. 3(a): Bundle rights protection with technology and resource interventions	X12 coefficient attenuates from marginal significance (Table 6) to non-significance in the integrated model (Table 7), confirming H2’s conditionality prediction	Table 6 and Table 7
Rec. 3(b): Low-friction-cost institutional design	Marginal significance of X12 (p = 0.0543) reflects its conditional, synergistic mechanism rather than autonomous operation	Table 6

maps each policy sub-recommendation to the specific empirical finding that motivates it.

6.3. Limitations and Future Directions

This study has several limitations that point toward productive directions for future research.

Sample scope and generalizability. The analytical sample is a full census of top-tier publishing and media enterprises on the Xinbang Index—not a probability draw from a wider population. Two implications follow. Statistically, the estimates are best read as descriptive and directional within this specific population; the population itself is small (n = 10), sector-specific, and country-specific. For external validity, China’s state-led platform governance, evolving copyright infrastructure, and concentrated market structure may together produce patterns that do not hold elsewhere. Future research should test whether the directional hierarchy (technology > rights > resource) replicates across multi-sector, multi-country, and multi-tier samples. Legislative variation deserves particular attention: copyright law, platform liability regimes, and digital competition rules differ substantially across jurisdictions and shape both the feasibility and the enforcement capacity of each regulatory dimension examined here. Incorporating legislative heterogeneity as an explicit moderator would test the framework’s boundary conditions and generate jurisdiction-sensitive guidance that the present study, grounded in China’s regulatory architecture, cannot provide.

Methodological constraints. This study relies on secondary platform analytics data. Combining regression analysis with Spearman rank correlation strengthens robustness within these data, but secondary metrics may miss how digital asset management shifts in fast-moving platform environments. Mixed-method designs—longitudinal panel data, platform-level trace data, or organizational case studies—would yield richer causal insight into how regulatory interventions produce equilibrium effects over time.

Measurement scope: environmental dimension. The Xinbang Index captures economic and social–institutional aspects of digital sustainability. It does not cover energy consumption, carbon emissions, or green technology adoption. The environmental sustainability claims in this paper are therefore confined to contextual framing (Section 2.6). Linking content–governance data with ecological performance metrics is a necessary step for future work addressing the environmental dimension directly.

Temporal scope: dynamic equilibrium. The cross-sectional design captures regulatory effects at one point in time. It cannot observe the equilibrium-restoring process itself (Section 3.1). The claim that the three dimensions steer the ecosystem toward dynamic equilibrium rests on observed directional complementarity—a structural inference, not a direct test. Confirming it requires longitudinal or panel data that track how platform ecosystems absorb and correct successive asymmetries. This is the most theoretically consequential direction for future research.

Demand-side information asymmetry. This study examines information asymmetry from the supply side, focusing on how regulatory interventions in resource allocation, technological innovation, and rights protection shape the distributional efficiency of content dissemination across platform actors. It does not account for heterogeneity on the demand side—namely, the variation among platform users in their capacity and willingness to acquire, process, and update information. Drawing on the concept of Bounded Rationality [28], even a supply-side symmetry-restoring intervention may leave residual asymmetries if users differ systematically in information literacy, attention allocation, or cognitive inertia. Future research should incorporate user-side factors into the analysis of regulatory effectiveness, examining whether and to what degree supply-side equilibrium-restoring efforts translate into genuine equilibrium outcomes at the user level.

Sector specificity. This study focuses on the publishing and media industry alone. The symmetry framework is theoretically generalizable, but its empirical base is sector-specific. Extending it to healthcare information platforms, educational technology ecosystems, or financial services platforms would both test transferability and reveal sector-specific moderators that shape the relative effectiveness of each governance dimension. A further constraint is data structure: several Xinbang Index variables exhibited near-zero cross-enterprise variance, narrowing the regression variable set. Firm-level survey data or proprietary platform analytics would enable richer variable specification and more granular hypothesis testing.