Informative Tweets Research Articles

Tracking the Spread of Pollen on Social Media Using Pollen-Related Messages From Twitter: Retrospective Analysis.

Allergy disorders caused by biological particles, such as the proteins in some airborne pollen grains, are currently considered one of the most common chronic diseases, and European Academy of Allergy and Clinical Immunology forecasts indicate that within 15 years 50% of Europeans will have some kind of allergy as a consequence of urbanization, industrialization, pollution, and climate change. The aim of this study was to monitor and analyze the dissemination of information about pollen symptoms from December 2006 to January 2022. By conducting a comprehensive evaluation of public comments and trends on Twitter, the research sought to provide valuable insights into the impact of pollen on sensitive individuals, ultimately enhancing our understanding of how pollen-related information spreads and its implications for public health awareness. Using a blend of large language models, dimensionality reduction, unsupervised clustering, and term frequency-inverse document frequency, alongside visual representations such as word clouds and semantic interaction graphs, our study analyzed Twitter data to uncover insights on respiratory allergies. This concise methodology enabled the extraction of significant themes and patterns, offering a deep dive into public knowledge and discussions surrounding respiratory allergies on Twitter. The months between March and August had the highest volume of messages. The percentage of patient tweets appeared to increase notably during the later years, and there was also a potential increase in the prevalence of symptoms, mainly in the morning hours, indicating a potential rise in pollen allergies and related discussions on social media. While pollen allergy is a global issue, specific sociocultural, political, and economic contexts mean that patients experience symptomatology at a localized level, needing appropriate localized responses. The interpretation of tweet information represents a valuable tool to take preventive measures to mitigate the impact of pollen allergy on sensitive patients to achieve equity in living conditions and enhance access to health information and services.

Sentiment Analysis Of Indonesian State Army Police Neutrality Sentiment Towards The 2024 Election On X Using The Support Vector Machine Algorithm

The accompanying goals are created: One method for figuring out the order of feeling examination in the balance of military police towards the 2024 political decision depends on popular assessment in SVM technique in arranging opinion investigation in the balance of military police towards the 2024 races in light of 2024 general assessment in X. In a leading examination, the stages utilized are the exploration system. This was finished to coordinate the exploration stages. The technique of this examination is quantitative. An exploration area is where a specialist completes research, particularly in catching peculiarities or examination that really happens at the exploration area to get precise and genuine examination information. The consequences of the testing did were to decide the capacity of the framework that had been made to complete feeling investigation on opinion towards the lack of bias of the TNI and Polri during the political race Research begins with compiling, specifically determining the points to be discussed. The subject of this research is the execution of message mining in testing the balance of military police feelings towards the 2024 political decisions in X using the Help vector machine1 algorithm. Tweet Information Collection,In this review, scientists utilized 800 tweet information.. The consequences of the opinion examination did will be introduced as a disarray framework, where through the disarray network and characterization report the degree of exactness of the exploration that has been completed can be determined.It is trusted that the aftereffects of this assessment can give a thorough image of the public's discernment on Twitter with respect to the lack of bias of the TNI and Polri in sorting out races.

Comparing location-specific and location-open social media data: methodological lessons from a study of blaming of minorities on Twitter during the COVID-19 pandemic

AbstractSocial media platforms such as Twitter (currently X) have become important sites of public discourse and participation. Researchers have attempted to identify and collect Twitter data within a certain country or region in order to answer research questions within a particular locale. However, location information of tweets is limited. Tackling the case of public blaming of minorities on Twitter in the context of the COVID-19 pandemic in the UK, we present a method for identifying UK-based tweets and analyse two types of datasets that we collected and processed: (a) tweets with UK location-tags (labelled as location-specific data and referred to as UK datasets); and (b) tweets with UK location-tags and / or user profiles containing potential UK location information (labelled as location-open data and referred to as ALL datasets). The empirical results reveal that the overall sentiments in the two dataset types align in the same direction, but the location-specific datasets contain more extreme discourses (i.e., more positive and more negative sentiments and fewer neutral sentiments). Furthermore, in the location-specific datasets, the range of theme areas is narrower, although the themes still grasp the essence of the discussion about blaming minorities found in the larger dataset. The findings demonstrate strengths and limitations of the two dataset types and that the location-specific data can be suitable especially when the available research resources are insufficient for collecting or processing larger datasets. Nevertheless, we propose that future research may consider comparing smaller and bigger datasets to test differences between these for other topics for which specific locations may be of particular interest.

Analyzing user reactions using relevance between location information of tweets and news articles

In this study, we analyze the extent of user reactions based on user’s tweets to news articles, demonstrating the potential for home location prediction. To achieve this, we quantify users’ reactions to specific news articles based on the textual similarity between tweets and news articles, showcasing that users’ reactions to news articles about their cities are significantly higher than those about other cities. To maximize the difference in reactions, we introduce the concept of News Distinctness, which highlights the news articles that affect a specific location. By incorporating News Distinctness with users’ reactions to the news, we magnify its effects. Through experiments conducted with tweets collected from users whose home locations are in five representative cities within the United States and news articles describing events occurring in those cities, we observed a 6.75% to 40% improvement in the reaction score when compared to the average reactions towards news for outside of home location, clearly predicting the home location. Furthermore, News Distinctness increases the difference in reaction score between news in the home location and the average of the news outside of the home location by 12% to 194%. These results demonstrate that our proposed idea can be utilized to predict the users’ location, potentially recommending meaningful information based on the users’ areas of interest.

A Deep Parallel Hybrid Fusion Model for disaster tweet classification on Twitter data

Disaster tweet classification has gained significant attention in natural language processing (NLP) due to its potential to aid disaster response and emergency management. The goal of disaster tweet classification is to automate the identification of informative tweets containing information related to various types of disasters, such as floods, earthquakes, wildfires, and more. This classification task plays a crucial role in real-time monitoring, situational awareness, and timely response coordination during emergency situations. In this context, we propose a Deep Parallel Hybrid Fusion Model (DPHFM) that combines features from Convolutional Neural Networks (CNNs) and Bidirectional Long Short-Term Memory (Bi-LSTM) as base learners. The features extracted from these base learners are combined using a fusion mechanism and then reconstructed for input to a meta-learner for making predictions. The DPHFM is trained on disaster datasets, such as crisisMMD, which consists of seven natural disaster events. The model underwent thorough evaluation using various metrics, demonstrating an average performance improvement of 90% to 96%. Furthermore, the proposed model’s performance surpassed that of other state-of-the-art models, showcasing its potential for disaster tweet classification using deep learning techniques.

Fake News Identification in Urdu Tweets Using Machine Learning Models

There is an increasing number of people who generate and distribute content online, especially via social media platforms, which is primarily responsible for the proliferation of fake information. Fake information can cause controversy and distort people's perspectives, so it needs to be addressed immediately. The goal of this work is to detect false information in Urdu tweets, a difficult task given the language's large user population and particular grammatical difficulties. We offer an all-inclusive machine learning system that reliably classifies tweets in Urdu as legitimate or false. The methodology we use consists of several key steps: preprocessing, which includes normalizing, tokenizing, removing stop words, and stemming to prepare the data for analysis; data collection, which involves compiling and annotating a sizable dataset of Urdu tweets; and feature extraction, which makes use of technique TF-IDF to extract the semantic and syntactic nuances of the language. We investigate various machine learning models, including RNNs and CNNs, and more sophisticated neural networks like SVM, Random Forest, Logistic Regression, Naive Bayes, and Decision Tree to find the most efficient method for resolving this classification problem. The models are put through a rigorous training and assessment process using measures including the F1 score, accuracy, precision, and recall. Furthermore, a thorough examination of their confusion matrices is done. Our study's findings suggest that deep learning models hold much promise for resolving the problem of inaccurate information in Urdu. This opens the door for additional research and the creation of real-time algorithms for spotting false information. The subject of information integrity in Urdu language content is improved by this work, which also sheds light on the applicability of machine learning techniques in many linguistic contexts. Using SVM, Random Forest, Logistic Regression, Naive Bayes, and Decision Tree we achieved accuracies of 74%,91%, 76%, 78%, and 67% respectively. Meanwhile, CNN and RNN are the classifiers with the highest accuracy levels at 91% and 99% respectively. The results demonstrate that the CNN Model achieved 99% highest accuracy in detecting fake news from Urdu Tweets.

Tweeting for change: social media narratives for sustainable service

PurposeThis interdisciplinary study aims to analyze how service organizations communicate sustainable beliefs in their social media narratives and use them to generate brand awareness, customer recognition and ongoing demand for sustainable service.Design/methodology/approachA two-phase exploratory analysis of 10,342 tweets from 2019–2020 was conducted by sustainable global corporations to identify best practices for their social media teams operating within a service-based business model. First, the significant themes were identified using an unguided machine learning approach of three types of firms: services, goods and mixed. Next, the full set of tweets with linguistic sentiment analysis was analyzed followed by a deeper view of the services-based organizations based on their strategic focus (business-to-business [B2B] versus mixed).FindingsThe findings indicate that tweets that appear to create the highest customer engagement are characterized as having high levels of analytical language, high clout (i.e. are socially relevant), a positive tone, a high number of words and a high number of words per sentence. On the other hand, having complex language in terms of six-letter words does not seem to associate with customer engagement. The last level of analysis shows that B2B services-based corporations with positive tone and higher word count exhibit higher levels of retweets. Implications include providing rational and informational tweets to increase engagement and highlight societal relevance.Originality/valueClimate change has negative consequences on human and physical capital, and ecosystems across the globe. This study provides specific recommendations for how services corporations can increase their sustainable communications and actions.Practical implicationsThe key implication of our research is that corporations must strategically design social media narratives about climate change as part of their online branding and communications process.

Mitigating healthcare supply chain challenges under disaster conditions: a holistic AI-based analysis of social media data

A key advantage of social media is the real-time exchange of views with large communities. In disaster situations, such bidirectional information exchange is most useful to victims and support teams, especially in communications with authorities, volunteers, and the public. This paper addresses challenges faced by the healthcare supply chain during the COVID-19 pandemic with analyses of Twitter data using an Artificial Intelligence-driven multi-step approach. We investigate tweets for information about healthcare supply chains, such as the scarcity of testing kits, oxygen cylinders, and hospital beds during the pandemic. We deployed machine learning to classify such tweets into imperative and non-imperative categories based on need severity. The study sought to predict the location of victims requesting help based on their imperative tweets if geo-tag information was missing. The proposed approach used four steps: (1) keyword-based informative tweet search, (2) raw tweet pre-processing, (3) content analysis to identify tweet trends, public sentiment, topics related to healthcare supply chain challenges, and crisis classification to label imperative and non-imperative tweets, (4) locating the point-of-crisis from imperative tweets to facilitate coordination of help operations. The pre-processing of tweets, trend analysis, and sentiment analysis relied on natural language processing and machine learning for topic modelling (K-mean clustering), crisis classification (random forest), and point-of-crisis detection (Markov chain). Results demonstrate the potential to capture significant, timely, and actionable information on healthcare supply chain challenges to respond quickly and appropriately in a pandemic.

Comparison and positioning of NGOs aimed at children from the perspective of social marketing on Twitter

The role of Non-Governmental Organizations (NGOs) in disseminating and protecting children’s rights is fundamental by increasing society’s knowledge about the reality that children face, thus mobilizing citizen attention. In this paper, we present an original study on social media data, specifically Twitter, to analyze childhood NGOs, evaluating the success of their content (through the likes obtained by publications) from the perspective of social marketing and prospective theory. In addition, it examines the positioning of organizations concerning the types of messages identified. The methodological approach is based on data mining, content analysis, and simple correspondence analysis through which the typology of the messages and positioning map are determined. The results suggest that these organizations generate predictable communication by publishing on specific topics and only increasing the number of tweets in emergencies when they are requiring urgent help. Some tweets show an immediate risk to which children are exposed if they do not receive help, which aligns with one of the premises of the Prospect Theory. Furthermore, a more significant number of posts does not necessarily imply a greater number of likes. Three types of messages have been determined: informative tweets that point out risks (type 1), impartial dialogue tweets (type 2), and action tweets that highlight benefits (type 3), confirmed through the Kruskal-Wallis test to have a relationship with impact. The positioning map shows that type 3 messages, which World Vision Spain opts for, are the most popular, followed by type 1, which Educo leans towards. Finally, there are those of type 2, with which UNICEF Spain is associated. The main implication is that our analysis validates the use of social media such as Twitter to analyze NGOs and proposes these social media platforms to be an important tool in mobilizing the community. In addition, this study offers parameters when constructing the messages for use in social marketing campaigns according to decisions that involve risk or certainty.

Sarcasm‐based tweet‐level stress detection

AbstractPsychological stress has evolved as an important health concern across the globe. The vulnerability to stress and the ramifications of it have only worsened during the time of the COVID‐19 pandemic. This necessitates a timely diagnosis of stress before the condition progresses to chronicity. In this context, the popularity of social media like Twitter, where large numbers of users share opinions without any social stigma, has emerged as a major resource of human opinions. This has led to an increased research interest in social media‐based stress detection techniques. However, tweet‐level stress detection techniques in the literature have left a void in leveraging the text information in tweets, especially the presence of sarcastic expressions in the tweet's text content. To this end, a novel method called “Sarcasm‐based Tweet‐Level Stress Detection” (STSD) is proposed in this work with the modification of the logistic loss function to detect tweet‐level stress by availing the information of sarcasm that exists in the tweet‐content. The principle of the STSD model is to minimize the loss for non‐sarcastic tweets while maximising the loss for sarcastic tweets. Furthermore, extensive preprocessing and dimensionality reduction are performed using kernel principal component analysis (kernel PCA) to improve the performance by reducing the dimensions. The experimental results show that the proposed STSD model, when applied along with kernel PCA, records a significant improvement in accuracy by a minimum of 5.25% and a maximum of 9.19% over baseline models. Also, there is an increment in F1‐score by at least 0.085 points and a maximum of 0.164 points when compared to the baseline models.

Функційно-стилістичні особливості неологізмів в українськомовному сегменті соціальної мережі "Твіттер"

The article outlines and analyzes the functional and stylistic parameters of neologisms in the Ukrainian-language segment of the social network "Twitter". The research data is based on more than two thousand Ukrainian-language Twitter posts in 2020-2024. Given the specifics of network, whose main purposes of which are providing communication and information, the following functions of neologisms in tweets are distinguished: nominative, communicative, attention- grabbing, influence-related, and pragmatic. The nominative function of the named words in tweets realizes the need to give new realities, individuals, phenomena, and events names that do not yet exist either in general usage or in lexicographical works. Neologisms also serve as a means of self-nomination and are part of aphorisms and paraphrases. The structure of tweets that represent the attention-grabbing function includes marker lexemes, with the help of which the author encourages online communicators to pay attention to a certain event, phenomenon, situation, or give their own assessment of it. Neologisms are used in interrogative statements, prompting communicators to hold their attention on posts, reflect on the information presented in them, and sometimes take the position of the tweet author in a reply. The cognitive function of neologisms is to verbalize the ability of the author of a tweet to understand, perceive, and learn information, i.e. to activate the author's speech and nominative activity with his or her pragmatic instructions. The analysis of the factual material suggests that since tweets are representatives of the author's views, and value judgments regarding certain events, concepts, and phenomena, it is advisable to distinguish the pragmatic function of neologisms in tweet expressions. The use of neologisms in tweets is influenced by pragmatic factors that reflect the author's attitude to the message; pragmatic information in tweets is expressed in evaluative neolexemes derived in different ways and represented mainly by noun and adjectival formations. The article focuses on analyzes neologisms with the morpheme ze-, which has recently become a productive class of neologisms.

Applying Text Mining and Semantic Network Analysis of Tweets to Investigate Social Perceptions on No Kids Zones in Korea

Objectives: This study aims to examine the social perceptions on the ‘No Kids Zones’ considering the continuous spread and the concern of the possibility of increased discrimination against children in our society.Methods: The study used a substantial number of tweets (<i>n</i> = 51,600) with the ‘No Kids Zones’ keyword. First, we probed tweets using Jupyter Notebook to create documents that contain texts between January 2019 and October 2022. Second, we examined TF-IDF, degree centrality, and frequency of words to extract the main keywords and reflect how relevant and connected those words are to a collection of tweet information. Third, based on the 60 keywords we extracted, CONCOR analysis and network visualization were performed using UCINET to examine the co-occurred perceptions on 'No Kids Zones'.Results: The findings demonstrate that the main keywords related to No Kids Zones include kid, parents, adult, children, baby, child and owner in the context of the social target, and ‘cafe’ as social place, and ‘discrimination’, ‘hate’ as social discourse. Second, the analysis showed that there are broad perspectives on children in the context of the No Kids Zones in Korean Society as being either difficult to handle (and, thus, causing inconvenience), or as deserving social understanding and consideration as growing beings.Conclusion: This study suggests a need for improvement both in social understanding and awareness of children and their rights, leading to the prevention of the spread of No Kids Zones. Moreover, parent-focused educational strategies to discipline their children are emphasized for the healthy development of children.

Find You: Multi-View-Based Location Inference for Twitter Users

Location information on Twitter plays a critical role in emergency detection, event recommendation, and disaster warning. However, only a limited amount of Twitter data are geotagged. Previous research has presented various models for inferring location based on text, social relations, and contextual data, yielding highly promising results. Nonetheless, these existing methods have certain limitations that need to be addressed. Firstly, most of the existing methods overlook the role of local celebrities (well-known users in the local community) as indicators of location within the social network. Secondly, they fail to consider the associations between words in tweets, resulting in insufficiently rich features extracted from the tweets. We propose a multi-view-based location inference model called MVGeo to overcome these limitations. In the network view, our approach employs the Gaussian Mixture Model (GMM) to identify and retain local celebrities, thereby strengthening user location associations. In the tweet view, we construct a heterogeneous graph based on the co-occurrence relationship between words in tweets and the user’s mentioned relationship with the words. This allows us to fully leverage the local correlation between words and the global correlation to extract tweet features more comprehensively. Finally, we employ a modified multi-layer graph convolutional network, called Gate-GCN, to fuse the network and tweet information. This expansion of the feature space enables us to extract sample features from multiple perspectives. To demonstrate the effectiveness of MVGeo, we conduct exhaustive experimental evaluations on a publicly available dataset and compare its performance against several state-of-the-art benchmark models. The results confirm the superior performance of the proposed model.

Green Tweets or Not? The Sustainable Commitment of Higher Education Institutions

Higher education organizations are increasingly playing a key role in promoting Sustainable Development within society by disclosing information about applied strategies and achieved sustainable performance via social media. Consequently, platforms like Twitter have evolved into decisive tools for fostering accountability, establishing legitimacy, and displaying engagement to environmental sustainability. First, this paper aims to analyze the impact of published sustainable content on Twitter on the stakeholders’ online engagement level within British green universities ranked by the People & Planet University League. Second, the study intends to investigate the influencing factors that enhance online interactions. To attain the purpose a theoretical framework was established and a text mining analysis of 145,941 tweets was conducted to construct both a sustainable and a general online engagement index. Finally, a multivariate regression analysis was employed to create two models based on content type. The results reveal a heightened interest in engaging with sustainable content shared on the official Twitter profiles of the sampled universities in comparison to general information tweets. Furthermore, the findings indicate that organizational attributes, reputational factors, and considerations related to strategy design and implementation of sustainable policies, along with the sentiment polarity of the disseminated content, contribute to elucidating the attained level of online engagement. These factors work together to involve stakeholders in the pursuit of sustainable development goals within Green Universities.

Comparing the Performance of Data Mining Algorithms in Predicting Sentiments on Twitter

On Twitter, users can post tweets, videos, and images. It can, however, also be disruptive and difficult. To categorize the material and improve searchability, hashtags are crucial. This study focuses on examining the opinions of Twitter users who participate in trending topics. The algorithms K-Nearest Neighbor (KNN) and Support Vector Machine (SVM) are used for sentiment analysis. The data set comprises tweet information on popular topics that was collected using the Twitter API and saved in Excel format. SVM and K-NN are used for data preparation, weighting, and sentiment analysis. With 105 data points, the study provides insight into user sentiment. SVM identified 99% of positive responses and 1% of negative responses with an accuracy of 80%. KNN successfully identified 90% of the positive responses and 10% of the negative responses, with an accuracy rate of 71.4%. According to the results, SVM performs better when analyzing the sentiment of hashtag users on Twitter.

Novel fuzzy deep learning approach for automated detection of useful COVID-19 tweets

Coronavirus (COVID-19) is a newly discovered viral disease from the SARS-CoV-2 family. This has caused a moral panic resulting in the spread of informative and uninformative information about COVID-19 and its effects. Twitter is a popular social media platform used extensively during the current outbreak. This paper aims to predict informative tweets related to COVID-19 on Twitter using a novel set of fuzzy rules involving deep learning techniques. This study focuses on identifying informative tweets during the pandemic to provide the public with trustworthy information and forecast how quickly diseases could spread. In this case, we have implemented RoBERTa and CT-BERT models using the fuzzy methodology to identify COVID-19 patient tweets. The proposed architecture combines deep learning transformer models RoBERTa and CT-BERT with the fuzzy technique to categorize posts as INFORMATIVE or UNINFORMATIVE. We performed a comparative analysis of our method with machine learning models and deep learning approaches. The results show that our proposed model can classify informative and uninformative tweets with an accuracy of 91.40% and an F1-score of 91.94% using the COVID-19 English tweet dataset. The proposed model is accurate and ready for real-world application.

What we tweet about when we tweet about taxes: A topic modelling approach

Recent literature on taxation suggests that a “service and client” approach by the authorities is required in order to establish a synergistic tax climate between taxpayers and tax offices and thus enhance voluntary tax compliance. The present study investigates whether lay people's conceptions about taxation reflect such a synergistic (vs. an antagonistic) climate. Applying an unsupervised machine learning approach (i.e., topic modeling) to over a million tax related tweets from 2010 to 2020, we identified 30 topics with different content. Using the theoretical framework differentiating between synergistic and antagonistic tax climate, we were able to further categorize these topics into four broader groups: 1. Opinions about Tax Politics, 2. Enforcement (antagonistic climate), 3. Information & Service (synergistic climate), and 4. Emotions. The most frequently observed group was Information & Service (synergistic climate), which also steadily gained prominence during the past decade. We proceeded by analyzing the information diffusion properties and sentiment of the tweets associated with the four groups. Information & Service tweets had the most positive sentiment but were shared the least, while tweets regarding Opinions about Tax Politics were shared most often. In sum, the results suggest that lay people's conceptions about taxation – as discerned from conversations on social media (Twitter) – largely reflect a synergistic (vs. an antagonistic) climate, and contribute to the literature on tax climate and social media.

A deep multi-view imbalanced learning approach for identifying informative COVID-19 tweets from social media

Social media platforms such as Twitter are home ground for rapid COVID-19-related information sharing over the Internet, thereby becoming the favorable data resource for many downstream applications. Due to the massive pile of COVID-19 tweets generated every day, it is significant that the machine-learning-supported downstream applications can effectively skip the uninformative tweets and only pick up the informative tweets for their further use. However, existing solutions do not specifically consider the negative effect caused by the imbalanced ratios between informative and uninformative tweets in training data. In particular, most of the existing solutions are dominated by single-view learning, neglecting the rich information from different views to facilitate learning. In this study, a novel deep imbalanced multi-view learning approach called D-SVM-2K is proposed to identify the informative COVID-19 tweets from social media. This approach is built upon the well-known multiview learning method SVM-2K to incorporate different views generated from different feature extraction techniques. To battle against the class imbalance problem and enhance its learning ability, D-SVM-2K stacks multiple SVM-2K base classifiers in a stacked deep structure where its base classifiers can learn from either the original training dataset or the shifted critical regions identified using the well-known k-nearest neighboring algorithm. D-SVM-2K also realises a global and local deep ensemble learning on the multiple views’ data. Our empirical experiments on a real-world labeled tweet dataset demonstrate the effectiveness of D-SVM-2K in dealing with the real-world multi-view class imbalance issues.

Analysis of perceptions and emotional repercussions in Twitter users in Colombia during the COVID-19 pandemic

IntroductionThe COVID-19 pandemic has negatively impacted mental health. Up to a quarter of the population has reported mental health disorders. This has been studied mainly from a nosological perspective, according to diagnostic criteria. Nevertheless, we did not find studies that have explored the daily expressions of the population. Our objective was to evaluate the perceptions of the COVID-19 pandemic and its repercussions on the emotional well-being of the Colombian population. MethodsWe performed a Twitter metrics and trend analysis. Initially, in the trend analysis, we calculated the average duration in hours of the 20 most popular trending topics of the day in Colombia and we grouped them into trends related to COVID-19 and unrelated trends. Subsequently, we identified dates of events associated with the pandemic relevant to the country, and they were related to the behaviour of the trends studied. Additionally, we did an exploratory analysis of these, selected the tweets with the greatest reach and categorised them in an inductive way to analyse them qualitatively. ResultsIssues not related to COVID-19 were more far-reaching than those related to coronavirus. However, a rise in these issues was seen on some dates consistent with important events in Colombia. We found expressions of approval and disapproval, solidarity and accusation. Inductively, we identified categories of informative tweets, humour, fear, stigma and discrimination, politics and entities, citizen complaints, and self-care and optimism. ConclusionsThe impact of the COVID-19 pandemic generates different reactions in the population, which increasingly have more tools to express themselves and know the opinions of others. Social networks play a fundamental role in the communication of the population, so this content could serve as a public health surveillance tool and a useful and accessible means of communication in the management of health crises.

Microblog topic detection using evolutionary clustering and social network information

SummaryThe messages in microblogs, like Twitter, present a wealth of information around the hot topics and public opinions. Topic detection and tracking enables a better understanding of the incoming information from social networks. Many of the studies performed in this field consider a set of predefined number of topics (clusters) which remain constant during the detection process. Such approaches are not efficient when data is dynamic and cumulative. In addition, nonparametric evolutionary topic models do not have an accurate performance on short texts due to the data deficiency. In this article, we propose an evolutionary clustering model based on the Bayesian nonparametric Dirichlet process mixture model, which automatically determines the number of clusters. To improve the algorithm performance when dealing with short social media texts, temporal information of tweets and other social network parameters are used in a weighted combination to augment the text similarity in detecting topics. Model evaluation is performed on the data gathered from the Twitter during a 2.5‐month period. Results reveal that the social network information can promote the topic detection performance. The proposed model outperforms previous study in terms of topic coherence and clustering performance, leading to a better clustering on short texts.