Search Results (95)

As software systems grow increasingly complex and interconnected, detecting vulnerabilities in source code has become a critical and challenging task. Traditional static analysis methods often fall short in capturing deep, context-dependent vulnerabilities and adapting to rapidly evolving threat landscapes. Recent efforts have explored knowledge graphs and transformer-based models to enhance semantic understanding; however, these solutions frequently rely on static knowledge bases, exhibit high computational overhead, and lack adaptability to emerging threats. To address these limitations, we propose DynaKG-NER++, a novel and lightweight framework for context-aware vulnerability detection in source code. Our approach integrates lexical, syntactic, and semantic features using a transformer-based token encoder, dynamic knowledge graph embeddings, and a Graph Attention Network (GAT). We further introduce contrastive learning on vulnerability–patch pairs to improve discriminative capacity and design an attention-based fusion module to combine token and entity representations adaptively. A key innovation of our method is the dynamic construction and continual update of the knowledge graph, allowing the model to incorporate newly published CVEs and evolving relationships without retraining. We evaluate DynaKG-NER++ on five benchmark datasets, demonstrating superior performance across span-level F1 (89.3%), token-level accuracy (93.2%), and AUC-ROC (0.936), while achieving the lowest false positive rate (5.1%) among state-of-the-art baselines. Sta tistical significance tests confirm that these improvements are robust and meaningful. Overall, DynaKG-NER++ establishes a new standard in vulnerability detection, balancing accuracy, adaptability, and efficiency, making it highly suitable for deployment in real-world static analysis pipelines and resource-constrained environments. Full article

The shift from syntagmatic to paradigmatic associations is a developmental process occurring from approximately the ages of six to nine years and plays an important role in language development. Syntagmatic relationships refer to words that co-occur due to their mutual dependency connection (e.g., “The dog barks”). Paradigmatic relationships are words within the same category (e.g., cat, kitten). In Study 1, we tested 244 Spanish-English bilingual children in grades 1 to 3 (M age = 7.87 years, 54.5% female) enrolled in dual language programs in California, USA. Children completed a matching task in both English and Spanish featuring both syntagmatic and paradigmatic lexical associations. Results showed significantly higher accuracy for older students than for younger students, higher accuracy in English than in Spanish for both paradigmatic and syntagmatic associations, and higher accuracy in paradigmatic associations in English and syntagmatic associations in Spanish. In Study 2, we conducted cognitive interviews with a separate sample of 13 Spanish-English bilingual children (M age = 8.96 years, 46.15% female) to explore how they reasoned through their word pair choices when completing the task. Children primarily relied on paradigmatic associations, using strategies like synonymy, antonymy, and category overlap, while also employing syntagmatic associations and thematic relatedness as less frequent but important reasoning strategies. Implications for early language development are discussed. Full article

In recent advancements within natural language processing (NLP), lexical networks play a crucial role in representing semantic relationships between words, enhancing applications from word sense disambiguation to educational tools. Traditional methods for constructing lexical networks, however, are resource-intensive, relying heavily on expert lexicographers. Leveraging GPT-4o, a large language model (LLM), our study presents an automated, scalable approach to creating multi-relational Japanese lexical networks for the general Japanese language. This study builds on previous methods of integrating synonyms but extends to other relations such as hyponymy, hypernymy, meronymy, and holonomy. Using a combination of structured prompts and graph-based data storage, the model extracts detailed lexical relationships, which are then systematically validated and encoded. Results reveal a substantial expansion in network size, with over 155,000 nodes and 700,000 edges, enriching Japanese lexical associations with nuanced hierarchical and associative layers. Comparisons with WordNet show substantial alignment in relation types, particularly with soft matching, underscoring the model’s efficacy in reflecting the multifaceted nature of lexical semantics. This work contributes a versatile framework for constructing expansive lexical resources that hold promises for enhancing NLP tasks and educational applications across various languages and domains. Full article

It is theoretically accepted that Chinese two-character words (2C-words) are processed both holistically and according to their constituent characters. Given the evidence on readers’ sensitivities to the syntactic relationships between the constituent characters, however, this general view might not fully explain the 2C-word processing mechanism. As an important category of 2C-words, verb–object constructions (VOCs) exhibit significant heterogeneity in splittability, the degree of syntactic phrasalization through the insertion of other characters between the constituent characters. To examine skilled readers’ VOC processing under the influences of splittability and whether the constituent characters are bound or free characters (character type), two experiments were conducted on a cohort of college students, who were Chinese native speakers, using the lexical decision task in a repetition priming paradigm. The prime stimuli (primer type) comprised three conditions: (a) the targets themselves, (b) the targets’ transposed non-words, and (c) non-linguistic baseline symbols ‘※※’. The primers’ two constituents were presented simultaneously and sequentially in Experiments 1 and 2, respectively. A significant interaction was revealed across both experiments between splittability and character type in the participants’ performance. The main effect was significant for primer type in the participants’ performance in Experiment 1; in Experiment 2, however, the interaction was significant both between primer type and splittability in the participants’ performance and between primer type and character type in their reaction times. In addition to confirming the general view, skilled readers might inevitably experience syntactic and semantic combinations of the constituent characters in their processing of VOCs. Full article

Students must navigate large catalogs of courses and make appropriate enrollment decisions in many online learning environments. In this context, identifying key concepts and their relationships is essential for understanding course content and informing course recommendations. However, identifying and extracting concepts can be an extremely labor-intensive and time-consuming task when it has to be done manually. Traditional NLP-based methods to extract relevant concepts from courses heavily rely on resource-intensive preparation of detailed course materials, thereby failing to minimize labor. As recent advances in large language models (LLMs) offer a promising alternative for automating concept identification and relationship inference, we thoroughly investigate the potential of LLMs in automatically generating course concepts and their relations. Specifically, we systematically evaluate three LLM variants (GPT-3.5, GPT-4o-mini, and GPT-4o) across three distinct educational tasks, which are concept generation, concept extraction, and relation identification, using six systematically designed prompt configurations that range from minimal context (course title only) to rich context (course description, seed concepts, and subtitles). We systematically assess model performance through extensive automated experiments using standard metrics (Precision, Recall, F1, and Accuracy) and human evaluation by four domain experts, providing a comprehensive analysis of how prompt design and model choice influence the quality and reliability of the generated concepts and their interrelations. Our results show that GPT-3.5 achieves the highest scores on quantitative metrics, whereas GPT-4o and GPT-4o-mini often generate concepts that are more educationally meaningful despite lexical divergence from the ground truth. Nevertheless, LLM outputs still require expert revision, and performance is sensitive to prompt complexity. Overall, our experiments demonstrate the viability of LLMs as a tool for supporting educational content selection and delivery. Full article

[Objective] Currently, there are limitations in the understanding of rural cultural landscape: they are often perceived as material spatial entities, with a lack of exploration of their intangible elements and neglect of the isomorphism between the material and intangible elements of cultural landscapes. In the context of rural cultural revitalization, it is necessary to explore the regional protection elements of rural cultural landscapes from the perspective of isomorphism. [Methods/Process] This study employs relevant linguistic theories to extract and construct a framework for a language system with regional characteristics for rural cultural landscapes from an isomorphous perspective. By deconstructing the rural cultural landscape pattern of Jiufangou in Dawu County, it summarizes the relationships and isomorphous nature between the constituent elements of this language system. [Results/Conclusions] The study identifies eight core landscape terms. These lexical units form landscape sentences based on four typical scenarios. The study then analyzed the landscape grammatical structures of different scenarios from four dimensions and explored the deep semantic meanings and contextual rules of Jiufanggou Village’s cultural landscape. Finally, this study utilizes a schematic diagram of the “vocabulary–grammar–sentence” nested structure of the Jiufanggou cultural landscape to visually illustrate the interconnections and patterns of cultural landscape elements in Jiufanggou Village across different contexts. Building on this, the study explores the structural equivalence between the material and immaterial elements of rural cultural landscapes. Overall, the construction of a nested linguistic system for rural cultural landscapes is not only about analyzing spatial forms but more importantly about exploring the underlying logical order and traditional wisdom behind spatial creation, thereby achieving the goals of associative protection, the inheritance of diverse cultures, and the continuation of the vitality of rural cultural landscapes. Full article

(1) Background: Perfectionism, generally conceptualized as a striving for flawlessness, can lead to maladaptive thoughts, feelings, and behavior. Both grandiose narcissism (GN) and vulnerable narcissism (VN) represent relevant personality dispositions for perfectionism. There is reason to assume that GN and VN predispose to different forms of perfectionist cognition and behavior. It remains unclear, however, whether GN and VN are indeed distinctly associated with different aspects of perfectionism and—if so—why. (2) Methods: We explored relationships between GN, VN, other-oriented, and socially prescribed perfectionism in a convenience sample of 210 adults (59% female) and further examined whether these relationships were mediated by distinct aspects of fear of failure, which has been identified as a critical driver for perfectionism. Moreover, we assessed implicit failure avoidance by means of response latencies obtained in a lexical approach-avoidance task. (3) Results: Our results indicate that perfectionist styles discriminate GN from VN whereby GN predict other-oriented and VN predict socially prescribed perfectionism. The latter relationship was largely mediated by social aspects of fear of failure (e.g., the fear of important others losing interest). In contrast, fear of failure did not explain the link between GN and other-oriented perfectionism. Furthermore, only VN was exclusively related to faster implicit failure avoidance. (4) Conclusions: This pattern of results suggests distinct mechanisms for GN and VN in the context of perfectionism. Our study provides support for the theoretical separation of GN and VN as relatively distinct phenotypes of narcissism and adds to clinical research linking GN and VN with different types of psychopathology. Full article

Late bilingual speakers immersed in a second language (L2) environment often experience the non-pathological attrition of their first language (L1), exhibiting selective and reversible changes in L1 processing and production. While attrition research has largely focused on long-term residents in anglophone countries, examining changes primarily within a single L1 domain, the present study employs a novel experimental design to investigate L1 attrition, alongside L2 acquisition, across three domains (i.e., the lexicon, syntax–pragmatics interface, and prosody) in two groups of L1-English L2-Italian late bilinguals: long-term residents in Italy vs. university students in the UK. A total of 112 participants completed online tasks assessing lexical retrieval, anaphora resolution, and sentence stress patterns in both languages. First, both bilingual groups showed comparable levels of semantic interference in lexical retrieval. Second, at the syntax–pragmatics interface, only residents in Italy showed signs of L1 attrition in real-time processing of anaphora, while resolution preferences were similar between groups; in the L2, both bilingual groups demonstrated target-like preferences, despite some slowdown in processing. Third, while both groups showed some evidence of target-like L2 prosody, with residents in Italy matching L1-Italian sentence stress patterns closely, prosodic attrition was only reported for residents in Italy in exploratory analyses. Overall, this study supports the notion of L1 attrition as a natural consequence of bilingualism—one that is domain- and experience-dependent, unfolds along a continuum, and involves a complex (and possibly inverse) relationship between L1 and L2 performance that warrants further investigation. Full article

Recent studies have greatly furthered our understanding of the Southern Bantu languages, but questions about the internal relationships of the Southern Bantu language subgroups and the validity of the clade as a whole still remain. This study attempts to reconstruct the consonant inventory of one proposed genetic clade, that of Tsonga-Copi (S50–S60). Using published dictionaries and reference works for each language of the subgrouping, a corpus of cognate vocabulary was assembled. Each term was then matched, where possible, to a reconstruction in the Bantu Lexical Reconstructions 3 (BLR3) database. Sound correspondences were identified and used to reconstruct the consonant inventory of Proto-Tsonga-Copi. In addition to the discovery of several typologically unusual sound changes, the results of this study also lend support to existing and developing hypotheses about both the internal relationships of Southern Bantu clades, as well as the nature of language contact in (pre)historic Southern Africa, particularly the influence of Khoisan and other Bantu languages. Full article

The pursuit of universal, symmetric semantic representations within large language models (LLMs) faces a fundamental challenge: the inherent asymmetry of natural languages. Different languages exhibit vast disparities in syntactic structures, lexical choices, and cultural nuances, making the creation of a truly shared, symmetric embedding space a non-trivial task. This paper aims to address this critical problem by introducing a novel framework to forge robust and symmetric multilingual sentence embeddings. Our approach, named DACL (Dynamic Asymmetric Contrastive Learning), is anchored in two powerful asymmetric learning paradigms: Contrastive Learning and Dynamic Curriculum Learning (DCL). We extend Contrastive Learning to the multilingual context, where it asymmetrically treats semantically equivalent sentences from different languages (positive pairs) and sentences with distinct meanings (negative pairs) to enforce semantic symmetry in the target embedding space. To further refine this process, we incorporate Dynamic Curriculum Learning, which introduces a second layer of asymmetry by dynamically scheduling training instances from easy to hard. This dual-asymmetric strategy enables the model to progressively master complex cross-lingual relationships, starting with more obvious semantic equivalences and advancing to subtler ones. Our comprehensive experiments on benchmark cross-lingual tasks, including sentence retrieval and cross-lingual classification (XNLI, PAWS-X, MLDoc, MARC), demonstrate that DACL significantly outperforms a wide range of established baselines. The results validate our dual-asymmetric framework as a highly effective approach for forging robust multilingual embeddings, particularly excelling in tasks involving complex linguistic asymmetries. Ultimately, this work contributes a novel dual-asymmetric learning framework that effectively leverages linguistic asymmetry to achieve robust semantic symmetry across languages. It offers valuable insights for developing more capable, fair, and interpretable multilingual LLMs, emphasizing that deliberately leveraging asymmetry in the learning process is a highly effective strategy. Full article

Background/Objectives: The study aimed to investigate how cognitive variables (performance IQ, verbal short-term memory, working memory, and ADHD symptomatology) impact lexical representations in children with high-functioning autism spectrum disorder (HF-ASD). Methods: Participants were two groups (n1 = n2 = 20) of monolingual Greek-speaking children, aged 7 to 12 years, with and without HF-ASD matched in age, gender, and cognitive skills. Results: Overall, the HF-ASD group had more immature lexical representations than the control group, even though the two groups were similar in naming. In both groups, naming was correlated moderately with verbal short-term memory but only age predicted significantly semantic knowledge. In the ASD group, a bilateral predictive relationship was revealed between output motor programming skills and stored phonological knowledge, supporting theoretical assumptions of the psycholinguistic model of speech. Finally, a different pattern of interrelations was observed between cognitive and lexical variables in the two groups. Conclusions: The findings of the current study indicate that ASD children may map and process new vocabulary differently compared to typically developing peers. Full article

Loess landslide disasters are influenced by a multitude of factors, including slope conditions, triggering mechanisms, and spatial attributes. Extracting these factors from unstructured geological texts is challenging due to nested entities, semantic ambiguity, and rare domain-specific terms. This study proposes a joint extraction framework guided by a domain ontology that categorizes six types of loess landslide influencing factors, including spatial relationships. The ontology facilitates conceptual classification and semi-automatic nested entity annotation, enabling the construction of a high-quality corpus with eight tag types. The model integrates a Soft-Lexicon mechanism that enhances character-level GloVe embeddings with explicit lexical features, including domain terms, part-of-speech tags, and word boundary indicators derived from a domain-specific lexicon. The resulting hybrid character-level representations are then fed into a BiLSTM-CRF architecture to jointly extract entities, attributes, and multi-level spatial and causal relationships. Extracted results are structured using a content-knowledge model to build a spatially enriched knowledge graph, supporting semantic queries and intelligent reasoning. Experimental results demonstrate improved performance over baseline methods, showcasing the framework’s effectiveness in geohazard information extraction and disaster risk analysis. Full article

Filled pauses are thought to be reflections of linguistic processes (e.g., lexical retrieval, speech planning and execution). Uh may be a self-directed cue for when a speaker needs more time to retrieve lexical–semantic representations, whereas um serves as a listener-directed, pragmatic cue. The use of filled pauses has not been examined in children who are hard of hearing (CHH). Participants included 68 CHH and 33 children with typical hearing (CTH). Participants engaged in conversations, expository discourse, and fable retells. We analyzed filled pauses as a function of hearing status and discourse contexts and evaluated the relationship between filled pauses and language ability. CHH produced uh across discourse contexts more often than their hearing peers. CHH did not differ in their use of um relative to CTH. Both um and uh were used more often in conversational samples compared to other types of discourse. Spearman’s correlations did not show any significant associations between the rate of filled pauses and standardized language scores. These results indicate that CHH produces uh more often than CTH, suggesting that they may have difficulty retrieving lexical–semantic items during ongoing speech. This information may be useful for interventionists who are collecting language samples during assessment. Full article

With the exponential growth of textual data, traditional topic modeling methods based on static analysis demonstrate limited effectiveness in tracking the dynamics of thematic content. This research aims to develop a method for quantifying the dynamics of topics within text corpora using a thematic signal (TS) function that accounts for temporal changes and semantic relationships. The proposed method combines associative tokens with original lexical units to reduce thematic entropy and information noise. Approaches employed include topic modeling (LDA), vector representations of texts (TF-IDF, Word2Vec), and time series analysis. The method was tested on a corpus of news texts (5000 documents). Results demonstrated robust identification of semantically meaningful thematic clusters. An inverse relationship was observed between the level of thematic significance and semantic diversity, confirming a reduction in entropy using the proposed method. This approach allows for quantifying topic dynamics, filtering noise, and determining the optimal number of clusters. Future applications include analyzing multilingual data and integration with neural network models. The method shows potential for monitoring information flows and predicting thematic trends. Full article

In social networks like community Question-Answering (cQA) services, members interact with each other by asking and answering each other’s questions. This way they find counsel and solutions to very specific real-life situations. Thus, it is safe to say that community fellows log into this kind of social network with the goal of satisfying information needs that cannot be readily resolved via traditional web searches. And in order to expedite this process, these platforms also allow registered, and many times unregistered, internauts to browse their archives. As a means of encouraging fruitful interactions, these websites need to be efficient when displaying contextualized/personalized material and when connecting unresolved questions to people willing to help. Here, demographic factors (i.e., gender) together with frontier deep neural networks have proved to be instrumental in adequately overcoming these challenges. In fact, current approaches have demonstrated that it is perfectly plausible to achieve high gender classification rates by inspecting profile images or textual interactions. This work advances this body of knowledge by leveraging lexicalized dependency paths to control the level of abstraction across texts. Our qualitative results suggest that cost-efficient approaches exploit distilled frontier deep architectures (i.e., DistillRoBERTa) and coarse-grained semantic information embodied in the first three levels of the respective dependency tree. Our outcomes also indicate that relative/prepositional clauses conveying geographical locations, relationships, and finance yield a marginal contribution when they show up deep in dependency trees. Full article