Sina Weibo Data Research Articles

Multivariate opinion evolution analysis on individual opinion differences: An agent-based model.

The spiral of silence is a classic theory originating from the age of mass media, which posits that individuals may suppress dissenting opinions to avoid social exclusion. In the digital era, despite the persistence of this phenomenon, the rise of social media has catalyzed the emergence of an antispiral of silence-a counterphenomenon fueled by growing self-awareness and online anonymity. In this paper, we focus on an interesting phenomenon: how these two spirals compete with each other to promote pluralism in social networks. To this end, we propose the spiral competition model (SCM) to quantify the two spirals and explore the impact of this phenomenon on the evolution of public opinion. In the SCM, agents can generate pseudo-opinions as well as antiopinions. In addition, we propose a time-varying network topology updating algorithm based on triangle closure, which simulates the fluid nature of social connections. Furthermore, we propose a community-based opinion initialization method with Gaussian distribution for enhancing simulation fidelity. Finally, we applied complex network theory to information dissemination, and the model was then simulated and analyzed in Erdős-Rényirandom networks. And the rationality of SCM was verified using Sina Weibo data. The experimental results show that while the spiral of silence initially holds sway, the antispiral of silence can precipitate significant opinion divergence during critical junctures, challenging the status quo. And if the antispiral of silence fails to prevail, the trend of public opinion will be difficult to reverse. This research helps us recognize the influence that can be brought about by the awakening of individual consciousness, and provides an important reference for public opinion prediction.

Evolution and spatiotemporal analysis of earthquake public opinion based on social media data

As critical conduits for the dissemination of online public opinion, social media platforms offer a timely and effective means for managing emergencies during major disasters, such as earthquakes. This study focuses on the analysis of online public opinions following the Maduo M7.4 earthquake in Qinghai Province and the Yangbi M6.4 earthquake in Yunnan Province. By collecting, cleaning, and organizing post-earthquake Sina Weibo (short for Weibo) data, we employed the Latent Dirichlet Allocation (LDA) model to extract information pertinent to public opinion on these earthquakes. This analysis included a comparison of the nature and temporal evolution of online public opinions related to both events. An emotion analysis, utilizing an emotion dictionary, categorized the emotional content of post-earthquake Weibo posts, facilitating a comparative study of the characteristics and temporal trends of online public emotions following the earthquakes. The findings were visualized using Geographic Information System (GIS) techniques. The analysis revealed certain commonalities in online public opinion following both earthquakes. Notably, the peak of online engagement occurred within the first 24 hours post-earthquake, with a rapid decline observed between 24 to 48 hours thereafter. The variation in popularity of online public opinion was linked to aftershock occurrences. Adjusted for population factors, online engagement in areas surrounding the earthquake sites and in Sichuan Province was significantly high. Initially dominated by feelings of “fear” and “surprise”, the public sentiment shifted towards a more positive outlook with the onset of rescue operations. However, distinctions in the online public response to each earthquake were also noted. Following the Yangbi earthquake, Yunnan Province reported the highest number of Weibo posts nationwide; in contrast, Qinghai Province ranked third post-Maduo earthquake, attributable to its smaller population size and extensive damage to communication infrastructure. This research offers a methodological approach for the analysis of online public opinion related to earthquakes, providing insights for the enhancement of post-disaster emergency management and public mental health support.

Self‐media information disclosure and firm productivity

AbstractUsing Sina Weibo data obtained through crawler technologies from 2013 to 2021, this study investigates the relationship between self‐media information disclosure and firm productivity. We find that self‐media information disclosure promotes firm productivity, a conclusion that withstands various robustness and endogeneity tests. Particularly, original microblogs exhibit a more significant impact. Further analysis reveals that the promoting effect is more pronounced in listed companies with lower annual retweets, small companies, and those with lower turnover. Finally, a mechanism analysis indicates that self‐media information disclosure enhances enterprise productivity by alleviating financing constraints, enhancing corporate innovation, and strengthening corporate governance.

Rally-around-the-destination? Changes in host-guest emotional solidarity after crises

Utilizing the political concept of “rally-around-the-flag”, this study takes Jiuzhai Valley as a case to construct a framework of host-guest emotional solidarity and examines the dynamics of such solidarity after an earthquake and the COVID-19 pandemic. Thematic and sentiment analyses are applied to interpret Sina Weibo data and the narratives derived from residents’ interviews as a form of triangulation. Results show that under crisis: (1) host-guest emotional solidarity includes shared beliefs, shared emotions, and interactive behaviors; (2) the short-term rally effect is emotion-driven and depends on “we-ness” relationships, emotional empathy, and place identity; (3) the long-term rally effect is value-driven and emerges when host-guest emotional empathy develops into value resonance through resource integration and relationship reconstruction. The results have implications for emotional solidarity theory and destination recovery in times of crisis.

Overview of big data information privacy protection technology and challenges

With the rapid development of advanced technologies like the mobile internet, cloud computing, and the Internet of Things, the era of big data is gradually upon us. In this era, big data has been widely utilized in various fields like business, finance, social networking, and healthcare, and the value of data has become increasingly prominent. Data has become a new factor of production, the digital economy has become a new engine of development, and an important driving force for promoting social progress and economic development. However, with the advancement of big data technology, the risks of data leakage and misuse have increased. Personal data information leakage or misuse can lead to personal privacy infringement and fraud, such as the Sina Weibo data leak in March 2020 where attackers breached part of the Sina Weibo database, obtaining sensitive data including real names, website usernames, genders, locations, and phone numbers, impacting 538 million Weibo users. Therefore, there is an increasing demand for data information privacy protection. This paper will explore the core technologies and application scenarios in data information privacy protection and investigate data privacy protection in real-life settings.

An insightful metric for evaluating perceived benefits from water quality enhancement in waterscape parks: A behavioral analysis approach

The rational evaluation of pollution control measures is crucial for advancing water pollution management, ensuring water quality, and creating high-standard urban environments. Currently the evaluation heavily relies on water quality indicators with little regard to the sense of gain of the people. In this paper, we proposed an objective indicator PVH (Park Visiting Hours) to represent the behavioral response, as a materialized sense of humans, to water quality. We demonstrated the application of PVH in four typical waterscape parks in Beijing with multi-source data, and the results showed that: (1) The Mann-Kendall Test and an independent correlation test utilizing Sina Weibo data (Chinese Twitter) indicated that PVH was significantly negatively correlated with the Satellite image based Sensory Pollution Index (SPI), as corroborated the competence of PVH to appraise the sense of gain from water quality improvement. (2) Multivariate regressions showed that an increase in SPI of 0.1 corresponds to an average 8.8 min reduction in PVH. (3) The relationship between PVH and SPI was consistent across various visitor demographics, and whether visitors were in groups or alone. This study introduced PVH as a novel but concise indicator to represent water quality-related behavior of park visitors and would inspire the use of big multisource data for measuring human perceptions of ecosystem services.

Topic and Trend Analysis of Weibo Discussions About COVID-19 Medications Before and After China's Exit from the Zero-COVID Policy: Retrospective Infoveillance Study.

After 3 years of its zero-COVID policy, China lifted its stringent pandemic control measures with the announcement of the 10 new measures on December 7, 2022. Existing estimates suggest 90%-97% of the total population was infected during December. This change created a massive demand for COVID-19 medications and treatments, either modern medicines or traditional Chinese medicine (TCM). This study aimed to explore (1) how China's exit from the zero-COVID policy impacted media and the public's attention to COVID-19 medications; (2) how social COVID-19 medication discussions were related to existing model estimates of daily cases during that period; (3) what the diversified themes and topics were and how they changed and developed from November 1 to December 31, 2022; and (4) which topics about COVID-19 medications were focused on by mainstream and self-media accounts during the exit. The answers to these questions could help us better understand the consequences of exit strategies and explore the utilities of Sina Weibo data for future infoveillance studies. Using a scrapper for data retrieval and the structural topic modeling (STM) algorithm for analysis, this study built 3 topic models (all data, before a policy change, and after a policy change) of relevant discussions on the Chinese social media platform Weibo. We compared topic distributions against existing estimates of daily cases and between models before and after the change. We also compared proportions of weibos published by mainstream versus self-media accounts over time on different topics. We found that Weibo discussions shifted sharply from concerns of social risks (case tracking, governmental regulations, etc) to those of personal risks (symptoms, purchases, etc) surrounding COVID-19 infection after the exit from the zero-COVID policy. Weibo topics of "symptom sharing" and "purchase and shortage" of modern medicines correlated more strongly with existing susceptible-exposed-infected-recovered (SEIR) model estimates compared to "TCM formulae" and other topics. During the exit, mainstream accounts showed efforts to specifically engage in topics related to worldwide pandemic control policy comparison and regulations about import and reimbursement of medications. The exit from the zero-COVID policy in China was accompanied by a sudden increase in social media discussions about COVID-19 medications, the demand for which substantially increased after the exit. A large proportion of Weibo discussions were emotional and expressed increased risk concerns over medication shortage, unavailability, and delay in delivery. Topic keywords showed that self-medication was sometimes practiced alone or with unprofessional help from others, while mainstream accounts also tried to provide certain medication instructions. Of the 16 topics identified in all 3 STM models, only "symptom sharing" and "purchase and shortage" showed a considerable correlation with SEIR model estimates of daily cases. Future studies could consider topic exploration before conducting predictive infoveillance analysis, even with narrowly defined search criteria with Weibo data.

BERT for sentiment analysis in the era of epidemic

With the continuous progress of Internet technology, the network platform has gradually entered everyone's life, providing a platform for ordinary people to express their ideas. Since the occurrence of COVID-19, monitoring and analyzing public opinion on the Internet platform has become more practical. Through timely monitoring and analysis, it is of great practical significance for the relevant departments to analyze and control sentiment information and stabilize and guide public sentiment. Therefore, it is essential and of practical significance to select a suitable model for classifying and analyzing public opinion on the Internet platform. This paper reviews the development of word vector technology from the perspective of technology development and then lead to the more advanced Bidirectional Encoder Representations from Transformers (BERT) model with great significance. On this basis, this paper fine-tunes the pre-trained Bert model. It applies the transfer learning strategy to analyzing the public sentiment of the occurrence of COVID-19 during the recent epidemic in Shanghai based on Sina Weibo data. In addition, tests are conducted to compare the model with the previous models. The experimental results show that the Bert model has significant advantages over the traditional model in character vector encoding and feature extraction.

Do stakeholders have the same concerns about anti-food waste law in China? Using big data from social media

As an important issue in global sustainable development, food waste and related waste management is an essential way to achieve China's “double carbon” goal and a critical factor in reaching global climate goals. To address the issue of food waste, the Chinese government issued the Anti-Food Waste Law (AFWL) in 2021. The implementation of the policy has had an impact on public life and sparked intense disputes on social media platforms. Based on Sina Weibo data, this study analyzed the connection between the main stakeholders and the AFWL from the multi-stage, multi-subject, and dynamic information perspective to clarify the current status of the AFWL implementation and then propose policy recommendations. The results show that public attention to AFWL is positively correlated with local economic development; the influence of integrated media accounts is more significant than that of specific food media accounts; stakeholders show a more consistent discourse tendency, but each has its focus according to the functions of each stakeholder. To promote anti-food waste efforts, the following policy recommendations are proposed: refine existing standards and develop localized policies, strengthen the guidance of opinion leaders and information monitoring in social media, set standards to regulate corporate food marketing models and promote digital marketing, and gradually cultivate consumers' awareness of food waste reduction. The study provides direction for optimizing AFWL in China, and the framework can be applied to other policy studies.

Research on the Characteristics of Internet Public Opinion and Public Sentiment after the Sichuan Earthquake Based on the Perspective of Weibo

In this paper, based on Sina Weibo data, a natural language processing (NLP) analysis method was used to analyze the temporal and spatial sequence characteristics of people’s attention and the characteristics of text content with the help of microblogs posted by people within 6 days after the 2022 Lushan M6.1, Maerkang M5.8 and Luding M6.8 earthquakes. Moreover, the same analysis method was used on the content of comments on microblogs posted by official media outlets within 6 days after the earthquakes to analyze the changes in people’s sentiments and the differences in the sentiments in various regions, and the influencing factors were also analyzed. The results of this research show the following: In terms of the spatial and temporal distributions, people’s attention was affected by the earthquakes themselves and their social impacts, and the first 2 h was often a period of an outbreak of attention, with the publishing areas mainly concentrated in Sichuan and Guangdong. In terms of people’s sentiments, the overall microblogging sentiment of the three earthquakes was positive, and the sentiment value of the people in Sichuan was generally low compared with that of the people in the other regions. Not only was the fluctuation in sentiment affected by the influence of the region, but it was also positively related to the sentiment of official microblogs. The results of this research provide reference for guiding people’s sentiments after earthquakes in the new media era.

Suicide Possibility Scale Detection via Sina Weibo Analytics: Preliminary Results.

Suicide, as an increasingly prominent social problem, has attracted widespread social attention in the mental health field. Traditional suicide clinical assessment and risk questionnaires lack timeliness and proactivity, and high-risk groups often conceal their intentions, which is not conducive to early suicide prevention. In this study, we used machine-learning algorithms to extract text features from Sina Weibo data and built a suicide risk-prediction model to predict four dimensions of the Suicide Possibility Scale-hopelessness, suicidal ideation, negative self-evaluation, and hostility-all with model validity of 0.34 or higher. Through this method, we can detect the symptoms of suicidal ideation in a more detailed way and improve the proactiveness and accuracy of suicide risk prevention and control.

A spatial-temporal topic model with sparse prior and RNN prior for bursty topic discovering in social networks

With the rapid growth of social network users, the social network has accumulated massive social network topics. However, due to the randomness of content, it becomes sparse and noisy, accompanied by many daily chats and meaningless topics, which brings challenges to bursty topics discovery. To deal with these problems, this paper proposes the spatial-temporal topic model with sparse prior and recurrent neural networks (RNN) prior for bursty topic discovering (ST-SRTM). The semantic relationship of words is learned through RNN to alleviate the sparsity. The spatial-temporal areas information is introduced to focus on bursty topics for further weakening the semantic sparsity of social network context. Besides, we introduced the “Spike and Slab” prior to decouple the sparseness and smoothness. Simultaneously, we realized the automatic discovery of social network bursts by introducing the burstiness of words as the prior and binary switching variables. We constructed multiple sets of comparative experiments to verify the performance of ST-SRTM by leveraging different evaluation indicators on real Sina Weibo data sets. The experimental results confirm the superiority of our ST-SRTM.

Assessing Reliability of Chinese Geotagged Social Media Data for Spatiotemporal Representation of Human Mobility

Understanding the space-time dynamics of human activities is essential in studying human security issues such as climate change impacts, pandemic spreading, or urban sustainability. Geotagged social media posts provide an open and space-time continuous data source with user locations which is convenient for studying human movement. However, the reliability of Chinese geotagged social media data for representing human mobility remains unclear. This study compares human movement data derived from the posts of Sina Weibo, one of the largest social media software in China, and that of Baidu Qianxi, a high-resolution human movement dataset from ‘Baidu Map’, a popular location-based service in China with 1.3 billion users. Correlation analysis was conducted from multiple dimensions of time periods (weekly and monthly), geographic scales (cities and provinces), and flow directions (inflow and outflow), and a case study on COVID-19 transmission was further explored with such data. The result shows that Sina Weibo data can reveal similar patterns as that of Baidu Qianxi, and that the correlation is higher at the provincial level than at the city level and higher at the monthly scale than at the weekly scale. The study also revealed spatial variations in the degree of similarity between the two sources. Findings from this study reveal the values and properties and spatiotemporal heterogeneity of human mobility data extracted from Weibo tweets, providing a reference for the proper use of social media posts as the data sources for human mobility studies.

Early detection of emergency events from social media: a new text clustering approach.

Emergency events require early detection, quick response, and accurate recovery. In the era of big data, social media users can be seen as social sensors to monitor real-time emergency events. This paper proposed an integrated approach to detect all four kinds of emergency events early, including natural disasters, man-made accidents, public health events, and social security events. First, the BERT-Att-BiLSTM model is used to detect emergency-related posts from massive and irrelevant data. Then, the 3 W attribute information (what, where, and when) of the emergency event is extracted. With the 3 W attribute information, we create an unsupervised dynamical event clustering algorithm based on text similarity and combine it with the supervised logistical regression model to cluster posts into different events. Experiments on Sina Weibo data demonstrate the superiority of the proposed framework. Case studies on some real emergency events show that the proposed framework has good performance and high timeliness. Practical applications of the framework are also discussed, followed by future directions for improvement.

Exploring spatiotemporal changes in the multi-granularity emotions of people in the city: a case study of Nanchang, China

In the Internet age, emotions exist in cyberspace and geospatial space, and social media is the mapping from geospatial space to cyberspace. However, most previous studies pay less attention to the multidimensional and spatiotemporal characteristics of emotion. We obtained 211,526 Sina Weibo data with geographic locations and trained an emotion classification model by combining the Bidirectional Encoder Representation from Transformers (BERT) model and a convolutional neural network to calculate the emotional tendency of each Weibo. Then, the topic of the hot spots in Nanchang City was detected through a word shift graph, and the temporal and spatial change characteristics of the Weibo emotions were analyzed at the grid-scale. The results of our research show that Weibo’s overall emotion tendencies are mainly positive. The spatial distribution of the urban emotions is extremely uneven, and the hot spots of a single emotion are mainly distributed around the city. In general, the intensity of the temporal and spatial changes in emotions in the cities is relatively high. Specifically, from day to night, the city exhibits a pattern of high in the east and low in the west. From working days to weekends, the model exhibits a low center and a four-week high. These results reveal the temporal and spatial distribution characteristics of the Weibo emotions in the city and provide auxiliary support for analyzing the happiness of residents in the city and guiding urban management and planning.

Culture vs. distance: comparing the effects of geographic segmentation variables on tourists\u2019 destination images based on social media data

Culture and distance are two major factors for geographically segmenting tourists in tourism marketing and advertising. Previous empirical studies on the destination image, however, have examined extensively the effect of the culture while inadequately the effect of the distance, let alone comparing the effects of the two variables. Using social media data, this study compares the effect of distance-based segments of tourists with that of culture-based segments in producing diverse perceived images of a destination. From Sina Weibo data, 282,532 Chinese mainland tourists who visited Suzhou, China during 2012–2016 and their perceived destination images are extracted and analyzed. The main results include: 1) for distance-based segments, the image differences increased with distance and the short-haul tourists perceived a more comprehensive image than the long-haul tourists; 2) for culture-based segments, the image differences were clear and relatively complex, while tourists from Wuyue cultural region had similar image perceptions with the local visitors; 3) the q-statistic of the Geodetector method shows that the culture-based segmentation can explain 65.8% of image variations while the distance-based segmentation can explain 46.6% of image variations, suggesting that culture is a more appropriate variable to segment the tourism market.

Perceiving Residents’ Festival Activities Based on Social Media Data: A Case Study in Beijing, China

Social media data contains real-time expressed information, including text and geographical location. As a new data source for crowd behavior research in the era of big data, it can reflect some aspects of the behavior of residents. In this study, a text classification model based on the BERT and Transformers framework was constructed, which was used to classify and extract more than 210,000 residents’ festival activities based on the 1.13 million Sina Weibo (Chinese “Twitter”) data collected from Beijing in 2019 data. On this basis, word frequency statistics, part-of-speech analysis, topic model, sentiment analysis and other methods were used to perceive different types of festival activities and quantitatively analyze the spatial differences of different types of festivals. The results show that traditional culture significantly influences residents’ festivals, reflecting residents’ motivation to participate in festivals and how residents participate in festivals and express their emotions. There are apparent spatial differences among residents in participating in festival activities. The main festival activities are distributed in the central area within the Fifth Ring Road in Beijing. In contrast, expressing feelings during the festival is mainly distributed outside the Fifth Ring Road in Beijing. The research integrates natural language processing technology, topic model analysis, spatial statistical analysis, and other technologies. It can also broaden the application field of social media data, especially text data, which provides a new research paradigm for studying residents’ festival activities and adds residents’ perception of the festival. The research results provide a basis for the design and management of the Chinese festival system.

Exploring the Spatiotemporal Patterns of Residents’ Daily Activities Using Text-Based Social Media Data: A Case Study of Beijing, China

The use of social media data provided powerful data support to reveal the spatiotemporal characteristics and mechanisms of human activity, as it integrated rich spatiotemporal and textual semantic information. However, previous research has not fully utilized its semantic and spatiotemporal information, due to its technical and algorithmic limitations. The efficiency of the deep mining of textual semantic resources was also low. In this research, a multi-classification of text model, based on natural language processing technology and the Bidirectional Encoder Representations from Transformers (BERT) framework is constructed. The residents’ activities in Beijing were then classified using the Sina Weibo data in 2019. The results showed that the accuracy of the classifications was more than 90%. The types and distribution of residents’ activities were closely related to the characteristics of the activities and holiday arrangements. From the perspective of a short timescale, the activity rhythm on weekends was delayed by one hour as compared to that on weekdays. There was a significant agglomeration of residents’ activities that presented a spatial co-location cluster pattern, but the proportion of balanced co-location cluster areas was small. The research demonstrated that location conditions, especially the microlocation condition (the distance to the nearest subway station), were the driving factors that affected the resident activity cluster patterns. In this research, the proposed framework integrates textual semantic analysis, statistical method, and spatial techniques, broadens the application areas of social media data, especially text data, and provides a new paradigm for the research of residents’ activities and spatiotemporal behavior.

EKGTF: A knowledge-enhanced model for optimizing social network-based meteorological briefings

With the frequent occurrence of extreme natural phenomena, news about meteorological disasters has increased. As a timely and effective social sensor, social networks have gradually become an important data source for the perception of extreme meteorological events. Meteorological briefing refers to screening valuable knowledge from massive data to provide decision-makers with efficient situational awareness support. However, social network-based briefing content has challenges, including colloquialisms and informal text styles. How to optimize these data in a formal text style is of great significance to improve decision-making efficiency. This paper proposes a meteorological briefing formalization module composed of three models: the text form judgment model, the formalization words detection model, and the event knowledge guided text formalization (EKGTF) model. These models are concatenated to optimize the meteorological briefing, specifically formalizing the briefing content’s language style based on Sina Weibo data. As a knowledge-enhanced model, the EKGTF model focuses on describing the core meteorological event knowledge while formalizing the content. Compared to baseline models, the EKGTF model achieves the best results on the BLEU score. Based on the meteorological briefing formalization module, a meteorological briefing formalization service framework is constructed, which is to be applied to the China Meteorological Administration (CMA) Public Meteorological Service Center.

Bidirectional Spatio-Temporal Association Between the Observed Results of Ulva Prolifera Green Tides in the Yellow Sea and the Social Response in Sina Weibo

Massive green tides caused by Ulva prolifera have annually occurred in the Yellow Sea since 2007, which has attracted much attention from the government and society. There are associations between the green tides in the Yellow Sea and social response in the social media (i.e., Sina Weibo), which are bidirectional and could be captured by the bidirectional neural network. For instance, how to detect daily U. prolifera green tides by fusing remote sensing data with social media data, and how to use the observed U. prolifera green tides to infer the social response are two challenges of State Oceanic Administration, China. This article first illustrated that there are bidirectional associations between green tides and Sina Weibo data. Then, this article introduced a bidirectional spatio-temporal associative memory neural network (BSAMNN) model for modeling this bidirectional association from the spatio-temporal perspective. BSAMNN first extracted six characteristics from green tides and nine characteristics from the social responses in 2016-2019. Second, these characteristics were split by year, and the characteristics in 2016-2018 were, respectively, put into the bidirectional associative memory neural network (BAM), which is a two-layer artificial neural network. Based on the BAM results and the observed data, the residual network was constructed. Third, BSAMNN extracted the spatio-temporal rules from the characteristics in 2016-2018 as the constraints through the mining rule algorithm and modify the results via sea surface wind and ocean surface current. Last, BSAMNN put the characteristics in 2019 into BAM and used the residual network to modify the results, which was constrained by the spatio-temporal rules. The feasibility and reliability of our approach were demonstrated by using the U. prolifera green tides in 2019. The average accuracy, false alarm rate, and missing alarm rate of BSAMNN results were 0.69, 0.25, and 0.31, respectively, which was 0.11 higher, 0.10 lower, and 0.11 lower than that of the traditional BAM. The results indicated that our method is an effective alternative of linking the U. prolifera green tides and its public sentiments on social media.