Search Results (15)

Once an online public opinion emerges, the interweaving of information distortion and public panic makes it difficult for enterprises to accurately grasp the emotional turning point and formulate sustainable marketing strategies. Based on the perspective of information asymmetry, in this paper, we construct a four-agent evolutionary game model involving the central government, local governments, enterprises and netizens. It analyzes the balance of strategies used by different actors in public opinion crises and examines how these strategies drive public panic from three perspectives: content, users and emotions. Finally, the findings are verified through simulation calculations. Our research reveals that when panic sentiment is in the medium range, the central government’s strengthened supervision coexists with enterprises’ deceptive marketing, and the impact of the event is magnified. When panic breaks through the threshold, local governments shift from full disclosure to partial disclosure, while consumers maintain their purchasing confidence and are less likely to be swayed by rumors. Research shows that after a public opinion crisis occurs, only by replacing deception with transparent and genuine content and jointly creating green solutions with consumers can enterprises transform panic into sustainable brand assets and provide a decision-making basis for the long-term development of the enterprise. Full article

Existing online review research has not fully captured consumer purchasing behavior in complex decision-making environments, particularly in contexts involving multiple product comparisons and conflicting review perspectives. This study thoroughly investigates the impact on focal product purchase decisions when consumers compare multiple products and face information inconsistency. Based on online review data from JD.com, we propose a semi-supervised deep learning model to analyze consumers’ sentiment polarity toward product attributes. The method establishes implicit relationships between labeled and unlabeled data through consistency regularization. Subsequently, we conceptualize three types of online review dissimilarity factors, rating-sentiment dissimilarity, cross-review dissimilarity, and brand dissimilarity, and employ regression models to examine the impact of competing products’ online reviews on focal product sales. The results indicate that by employing a semi-supervised deep learning approach, unlabeled data are annotated with pseudo-labels and utilized for model training, achieving more accurate sentiment classification than using labeled data alone. Moreover, positive (negative) sentiment attributes of competing products have a significant negative (positive) effect on focal product purchases. Online review dissimilarity moderates the spillover effects of competing products. Notably, these spillover effects are more pronounced when competing products are from the same brand compared to different brands. The research findings not only highlight the heterogeneous effects of positive and negative sentiments but also provide a new perspective for examining dissimilarity, enriching the understanding of online review spillover effects and the role of dissimilarity, while offering practical guidance for resource allocation decisions by companies and platforms. Full article

In the research into semantic sentiment analysis, researchers commonly use some factor rules, such as the utilization of emotional keywords and the manual definition of emotional rules, to increase accuracy. However, this approach often requires extensive data and time-consuming training, and there is a need to make the system simpler and more efficient. Recognizing these challenges, our paper introduces a new semantic sentiment analysis system designed to be both higher in quality and more efficient. The structure of our proposed system is organized into several key phases. Initially, we focus on data training, which involves studying emotions and emotional psychology. Utilizing linguistic resources such as HowNet and the Chinese Knowledge and Information Processing (CKIP) techniques, we develop emotional rules that facilitate the generation of sparse representation characteristics. This process also includes constructing a sparse representation dictionary. We can map these back to the original vector space by resolving the sparse coefficients, representing two distinct categories. The system then calculates the error compared to the original vector, and the category with the minimum error is determined. The second phase involves inputting topics and collecting relevant comments from internet forums to gather public opinion on trending topics. The final phase is data classification, where we assess the accuracy of classified issues based on our data training results. Additionally, our experimental results will demonstrate the system’s ability to identify hot topics, thus validating our semantic classification models. This comprehensive approach ensures a more streamlined and effective system for semantic sentiment analysis. Full article

Manually tuning the hyperparameters of a deep learning model is not only a time-consuming and labor-intensive process, but it can also easily lead to issues like overfitting or underfitting, hindering the model’s full convergence. To address this challenge, we present a BiLSTM-TCSA model (BiLSTM combine TextCNN and Self-Attention) for deep learning-based sentiment analysis of short texts, utilizing an improved particle swarm optimization (IPSO). This approach mimics the global random search behavior observed in bird foraging, allowing for adaptive optimization of model hyperparameters. In this methodology, an initial step involves employing a Generative Adversarial Network (GAN) mechanism to generate a substantial corpus of perturbed text, augmenting the model’s resilience to disturbances. Subsequently, global semantic insights are extracted through Bidirectional Long Short Term Memory networks (BiLSTM) processing. Leveraging Convolutional Neural Networks for Text (TextCNN) with diverse convolution kernel sizes enables the extraction of localized features, which are then concatenated to construct multi-scale feature vectors. Concluding the process, feature vector refinement and the classification task are accomplished through the integration of Self-Attention and Softmax layers. Empirical results underscore the effectiveness of the proposed approach in sentiment analysis tasks involving succinct texts containing limited information. Across four distinct datasets, our method attains impressive accuracy rates of 91.38%, 91.74%, 85.49%, and 94.59%, respectively. This performance constitutes a notable advancement when compared against conventional deep learning models and baseline approaches. Full article

Government may need to launch policies to stabilize real estate prices being away from unusual rise at an unexpected pace through short-term regulations of sales and purchases. Short-term control policies are often not effective immediately after withdrawal, but their effect easily attracts swift and intensive responses of consumer sentiments. The change in sentiment synchronizes with that of expectations, which together account for housing price in response to restrictions following short-term policies. The research objective of this study is to establish the role of housing sentiment in policymaking to regulate and stabilize real estate prices. To cope with the tough tissue of unclear knowledge about customers’ sentiments, we employed the state-space model to explore the impact of short-term regulatory policies on housing sentiment. The research objective of this study also involves optimizing the instrument for assessing housing sentiments. Results showed that: Firstly, the short-term regulation and control policy enhanced positive sentiment in the housing market. Secondly, high positive sentiment further increased the cyclical prices. Thirdly, the upsurge of consumer sentiment has weakened the impact of short-term control policies on real estate market price. Lowered housing sentiment resulted in a reduction in the effectiveness of short-term control policies. Overall, our study verifies that high positive consumer sentiments will result in an increase in housing prices, hence it is customers’ sentiments that caused the failure of short-term control policies. Full article

This study mapped personality based on the newly proposed extraction method from consumers’ textual data and revealed the relevance (attention) and polarity (affection) of words associated with a specific personality trait. Furthermore, we illustrate how unique words are used to predict a consumer’s behavior associated with certain personality traits. In this study, we employed the scales of the Kaggle MBTI Personality dataset to examine the methodology’s effectiveness, extract the personality traits from the textual data into features, and map them into the traits/dimensions of the existing scale. Based on the results obtained in this study, we assert that using the TF-IDF algorithm is a good way to generate a custom dictionary. Furthermore, sentiment scoring with an AI-empowered machine learning algorithm provides useful data to filter and validate more coherent words to understand and, thus, communicate a particular aspect of personality. Finally, we proposed that four situations involving the interaction between attention (frequency) and affection (sentiment) allow us to better understand the consumer and how to use the feature words in terms of the interaction between attention (TF-IDF score) and affection (sentiment score). Full article

Growing Mediterranean seafood consumption, increasing consumers’ awareness of food safety and quality, and changing food lifestyles are leading to the development of new food products. However, the majority of new food products launched on the market are expected to fail within the first year. One of the most effective ways to enhance new product success is by involving consumers during the first phases of New Product Development (NPD), using the so-called co-creation approach. Based on data collected through online discussion rooms, two new seafood product concepts—sardine fillets and sea burgers—were evaluated by a set of potential consumers in three Mediterranean countries—Italy, Spain, and Croatia. Textual information was analyzed by first using the topic modeling technique. Then, for each main topic identified, sentiment scores were calculated, followed by the identification of the main related emotions that were evoked. Overall, consumers seem to positively evaluate both proposed seafood product concepts, and three recurrent positive emotions (trust, anticipation, joy) were identified in relation to the main topics aroused during the discussions. The results of this study will be useful to guide future researchers and actors in this industry in the next development steps of the targeted seafood products in Mediterranean countries. Full article

In the Web2.0 era, user-generated content (UGC) provides a valuable source of data to aid in understanding consumers and driving intelligent business. Text mining techniques, such as semantic analysis and sentiment analysis, help to extract meaningful information embedded in UGC. However, research on text mining of UGC for e-commerce business applications involves interdisciplinary knowledge, and few studies have systematically summarized the research framework and application directions of related research in this field. First, based on e-commerce practice, in this study, we derive a general framework to summarize the mainstream research in this field. Second, widely used text mining techniques are introduced, including semantic and sentiment analysis. Furthermore, we analyze the development status of semantic analysis in terms of text representation and semantic understanding. Then, the definition, development, and technical classification of sentiment analysis techniques are introduced. Third, we discuss mainstream directions of text mining for business applications, ranging from high-quality UGC detection and consumer profiling, to product enhancement and marketing. Finally, research gaps with respect to these efforts are emphasized, and suggestions are provided for future work. We also provide prospective directions for future research. Full article

Research into news media representations of Muslims and their faith has focused mainly on how Muslims are portrayed in various types of news media and how stories about or involving them are framed. However, there has been very little attention paid to the effects of news consumption on attitudes towards Muslims. Accordingly, we wanted to explore a range of issues associated with news consumption levels and attitudes towards Muslims in Australia. The three objectives of this article are to: explore whether the amount of news consumed by respondents to an Australian survey influences the level of animosity they hold towards Muslims; determine how political viewpoint and religiosity influence the relationship between news consumption and animosity towards Muslims; and see whether engagement with Muslims influences the relationship between news consumption and animosity towards Muslims. Through a 2018 nationally representative sample of Australians, we target these objectives by investigating whether the amount of news that non-Muslim survey participants consume in a week influences the levels of anger they feel towards Muslims and how their self-defined religiosity, political viewpoint, and engagement with Muslims affect that relationship, while controlling for known drivers of anti-Muslim sentiment, such as demographic characteristics and knowledge about Muslims. We set our study in the contemporary context of mostly lab-based research that helps us understand how news media consumption affects particular types of people and whether there are commonalities in like-groups’ responses to different types of news consumption; in this case, stories about Muslims and their faith. The findings of our research will be of interest to news media organizations and journalists wanting to know about the effects of their coverage of stories about Muslims and their faith and those wanting to improve that reportage. The results will also interest groups working on social cohesion efforts, those trying to improve inter-faith and inter-cultural relations, and academics investigating news media coverage of Muslims and Islam. Significantly, we find quantity of news consumption to lack effect on anger levels. Full article

Internet service providers (ISPs) conduct their business by providing Internet access features to their customers. The COVID-19 pandemic has shifted most activity being performed remotely using an Internet connection. As a result, the demand for Internet services increased by 50%. This significant rise in the appeal of Internet services needs to be overtaken by a notable increase in the service quality provided by ISPs. Service quality plays a great role for enterprises, including ISPs, in retaining consumer loyalty. Thus, modelling ISPs’ service quality is of great importance. Since a common technique to reveal service quality is a timely and costly pencil survey-based method, this work proposes a framework based on the Sentiment Analysis (SA) of the Twitter dataset to model service quality. The SA involves the majority voting of three machine learning algorithms namely Naïve Bayes, Multinomial Naïve Bayes and Bernoulli Naïve Bayes. Making use of Thaicon’s service quality metrics, this work proposes a formula to generate a rating of service quality accordingly. For the case studies, we examined two ISPs in Indonesia, i.e., By.U and MPWR. The framework successfully extracted the service quality rate of both ISPs, revealing that By.U is better in terms of service quality, as indicated by a service quality rate of 0.71. Meanwhile, MPWR outperforms By.U in terms of customer service. Full article

With the development of Internet technology and the digital market in China, consumption through e-commerce platforms and giving online reviews after purchase have become mainstream. However, current market research still uses traditional methods of surveys and questionnaires, which largely influences the efficiency of relevant enterprises in finding key product issues and lowers new product performance. This paper conducts the aspect-based sentiment analysis (ABSA) method to study the impact of consumer sentiment involvement (CSI) on new product performance. We took the ceramic industry as a case study, and collected 3.22 million consumer responses for ABSA. A total of 22 performances of new products were analyzed for CSI hypothesis testing. We found that CSI big data analytics were positively related to new product performance and enterprise innovation. Our study contributes in three ways. First, we extend the concept of co-creation in open innovation studies into an intelligent data intensive context. The data-driven open innovation is conducive to product design and new product performance. Second, we enrich the empirical industry of ABSA sentiment analysis in the Chinese ceramic industry. Third, we contribute to the collaboration method to motivate consumers to participate, which means to make effective product reviews in this research. Full article

Recommender systems are being used in streaming service platforms to provide users with personalized suggestions to increase user satisfaction. These recommendations are primarily based on data about the interaction of users with the system; however, other information from the large amounts of media data can be exploited to improve their reliability. In the case of media social data, sentiment analysis of the opinions expressed by users, together with properties of the items they consume, can help gain a better understanding of their preferences. In this study, we present a recommendation approach that integrates sentiment analysis and genre-based similarity in collaborative filtering methods. The proposal involves the use of BERT for genre preprocessing and feature extraction, as well as hybrid deep learning models, for sentiment analysis of user reviews. The approach was evaluated on popular public movie datasets. The experimental results show that the proposed approach significantly improves the recommender system performance. Full article

The association between changes in public sentiment induced by COVID-19 and green consumption behavior has not been studied deeply. This study proposes that the awe induced by the COVID-19 pandemic can have both negative and positive aspects, aiming to psychologically reveal why the pandemic is affecting green consumer behavior and explore potential pathways for differentiation. Research data were derived from Wuhan, China, and analyzed using experimental method. This study finds that awe of COVID-19 positively affects green consumption behavior. Specifically, due to fear, anxiety, and powerlessness, individuals with negative awe of COVID-19 instinctively need to respond to risk and pay more attention to their own safety and interests, so as to promote green consumption. However, positive awe of COVID-19 involves higher levels of cognition, such as admiration, inspiration, and optimism. It inspires a commitment to prioritize nature and social groups, and promotes green consumption behavior. As conclusions, different types of awe can be induced from public health emergencies like COVID-19 and have their own specific paths to effect green consumption behavior. These findings could help governments and marketers build future policies and strategies to reasonably guide public sentiment in order to better promote green consumption in this epidemic. Full article

The growth of ecommerce has triggered online reviews as a rich source of product information. Revealing consumer sentiment from the reviews through Sentiment Analysis (SA) is an important task of online product review analysis. Two popular approaches of SA are the supervised approach and the lexicon-based approach. In supervised approach, the employed machine learning (ML) algorithm is not the only one to influence the results of SA. The utilized text features also handle an important role in determining the performance of SA tasks. In this regard, we proposed a method to extract text features that takes into account semantic of words. We argue that this semantic feature is capable of augmenting the results of supervised SA tasks compared to commonly utilized features, i.e., bag-of-words (BoW). To extract the features, we assigned the correct sense of the word in reviewing the sentence by adopting a Word Sense Disambiguation (WSD) technique. Several WordNet similarity algorithms were involved, and correct sentiment values were assigned to words. Accordingly, we generated text features for product review documents. To evaluate the performance of our text features in the supervised approach, we conducted experiments using several ML algorithms and feature selection methods. The results of the experiments using 10-fold cross-validation indicated that our proposed semantic features favorably increased the performance of SA by 10.9%, 9.2%, and 10.6% of precision, recall, and F-Measure, respectively, compared with baseline methods. Full article

People increasingly can and want to obtain and generate health information themselves. With the increasing do-it-yourself sentiment comes also the desire to be more involved in one’s health care decisions. Patient driven health-care and health research models are emerging; terms such as participatory medicine and quantified-self are visible increasingly. Given the health consumer’s desire to be more involved in health data generation and health care decision making processes the authors submit that it is important to be health policy literate, to understanding how health policies are developed, what themes are discussed among health policy researchers and policy makers, to understand how ones demands would be discussed within health policy discourses. The public increasingly obtains their knowledge through the internet by searching web browsers for keywords. Question is whether the “health consumer” to come has knowledge of key terms defining key health policy discourses which would enable them to perform targeted searches for health policy literature relevant to their situation. The authors found that key health policy terms are virtually absent from printed and online news media which begs the question how the “health consumer” might learn about key health policy terms needed for web based searches that would allow the “health consumer” to access health policy discourses relevant to them. Full article