Showing papers on "Microblogging published in 2017"
19 Oct 2017
TL;DR: A novel Recurrent Neural Network with an attention mechanism (att-RNN) to fuse multimodal features for effective rumor detection and the results demonstrate the effectiveness of the proposed end-to-end att- RNN in detecting rumors with multi-modal contents.
Abstract: Microblogs have become popular media for news propagation in recent years. Meanwhile, numerous rumors and fake news also bloom and spread wildly on the open social media platforms. Without verification, they could seriously jeopardize the credibility of microblogs. We observe that an increasing number of users are using images and videos to post news in addition to texts. Tweets or microblogs are commonly composed of text, image and social context. In this paper, we propose a novel Recurrent Neural Network with an attention mechanism (att-RNN) to fuse multimodal features for effective rumor detection. In this end-to-end network, image features are incorporated into the joint features of text and social context, which are obtained with an LSTM (Long-Short Term Memory) network, to produce a reliable fused classification. The neural attention from the outputs of the LSTM is utilized when fusing with the visual features. Extensive experiments are conducted on two multimedia rumor datasets collected from Weibo and Twitter. The results demonstrate the effectiveness of the proposed end-to-end att-RNN in detecting rumors with multimodal contents.
374 citations
TL;DR: This paper explores the key role of image content in the task of automatic news verification on microblogs and proposes several visual and statistical features to characterize these patterns visually and statistically for detecting fake news.
Abstract: Microblog has been a popular media platform for reporting and propagating news. However, fake news spreading on microblogs would severely jeopardize its public credibility. To identify the truthfulness of news on microblogs, images are very crucial content. In this paper, we explore the key role of image content in the task of automatic news verification on microblogs. Existing approaches to news verification depend on features extracted mainly from the text content of news tweets, while image features for news verification are often ignored. According to our study, however, images are very popular and have a great influence on microblogs news propagation. In addition, fake and real news events have different image distribution patterns. Therefore, we propose several visual and statistical features to characterize these patterns visually and statistically for detecting fake news. Experiments on a real-world multimedia dataset collected from Sina Weibo validate the effectiveness of our proposed image features. The news verification performance of our method outperforms baseline methods. To the best of our knowledge, this is the first attempt that systematically explores image features on news verification task.
323 citations
TL;DR: It was found that Twitter sentiment and posting volume were relevant for the forecasting of returns of S&P 500 index, portfolios of lower market capitalization and some industries, and KF sentiment was informative for the forecast of returns.
Abstract: In this paper, we propose a robust methodology to assess the value of microblogging data to forecast stock market variables: returns, volatility and trading volume of diverse indices and portfolios. The methodology uses sentiment and attention indicators extracted from microblogs (a large Twitter dataset is adopted) and survey indices (AAII and II, USMC and Sentix), diverse forms to daily aggregate these indicators, usage of a Kalman Filter to merge microblog and survey sources, a realistic rolling windows evaluation, several Machine Learning methods and the Diebold-Mariano test to validate if the sentiment and attention based predictions are valuable when compared with an autoregressive baseline. We found that Twitter sentiment and posting volume were relevant for the forecasting of returns of S&P 500 index, portfolios of lower market capitalization and some industries. Additionally, KF sentiment was informative for the forecasting of returns. Moreover, Twitter and KF sentiment indicators were useful for the prediction of some survey sentiment indicators. These results confirm the usefulness of microblogging data for financial expert systems, allowing to predict stock market behavior and providing a valuable alternative for existing survey measures with advantages (e.g., fast and cheap creation, daily frequency).
255 citations
TL;DR: The anatomy of the information space on Facebook is explored by characterizing on a global scale the news consumption patterns of 376 million users over a time span of 6 y, finding that users tend to focus on a limited set of pages, producing a sharp community structure among news outlets.
Abstract: The advent of social media and microblogging platforms has radically changed the way we consume information and form opinions. In this paper, we explore the anatomy of the information space on Facebook by characterizing on a global scale the news consumption patterns of 376 million users over a time span of 6 y (January 2010 to December 2015). We find that users tend to focus on a limited set of pages, producing a sharp community structure among news outlets. We also find that the preferences of users and news providers differ. By tracking how Facebook pages “like” each other and examining their geolocation, we find that news providers are more geographically confined than users. We devise a simple model of selective exposure that reproduces the observed connectivity patterns.
225 citations
TL;DR: The findings reveal that Twitter was most commonly used for communication and assessment purposes, and proposed five guidelines that could help promote the educational value of Twitter use.
Abstract: Twitter, a popular microblogging social networking site, allows individuals to communicate by sending short messages of up to 140 characters. Although it enables people to be in constant contact, its value in educational context is less clear. This paper is the first to examine empirical studies of using Twitter in teaching and learning over 10 years from 2006 to 2015, with the aim of understanding whether its implementation would benefit students or not. We identified a total of 51 eligible publications, and reported the analysis in four major categories: (a) the profile of studies, (b) the specific ways in which Twitter was employed in education, (c) the impacts on interactions, and (d) the impacts on students' learning outcomes. The findings reveal that Twitter was most commonly used for communication and assessment purposes. Although Twitter shows promise in improving interactions among learners and teachers, causality between Twitter use and learning performance remains to be conclusively established. Currently, the most beneficial use of Twitter is probably that of a “push” technology – such as the instructor sending important course information, homework assignments and test deadlines to students, as well as that of a platform for peer interaction. Many challenges still exist in using Twitter for teaching and learning. Based on our review of the literature, we proposed five guidelines that could help promote the educational value of Twitter use. We also identified several limitations of previous studies, and offered suggestions for future work.
181 citations
TL;DR: In this article, a theoretical framework of the influence of microblogs and consumer reviews on new product success has been proposed, drawing from consumer information search theory and diffusion theory, which analyzes a system of equations.
110 citations
TL;DR: The findings indicate that the level, length, type and attitude of retailers' engagement with social media users have a significant impact on their sentiments, which associate with brand image, perception and customer service of the online retailers.
97 citations
TL;DR: Wang et al. as discussed by the authors used a randomized field experiment on Sina Weibo, the top tweeting website in China, to examine how the viewing of TV shows is affected by the media company's tweets about its shows, and recruited Weibo influentials' retweets of the company tweets.
Abstract: Many businesses today have adopted tweeting as a new form of product marketing. However, whether and how tweeting affects product demand remains inconclusive. The authors explore this question using a randomized field experiment on Sina Weibo, the top tweeting website in China. The authors collaborate with a major global media company and examine how the viewing of its TV shows is affected by (1) the media company’s tweets about its shows, and (2) recruited Weibo influentials’ retweets of the company tweets. The authors find that both company tweets and influential retweets increase show viewing, but in different ways. Company tweets directly boost viewing, whereas influential retweets increase viewing if the show tweet is informative. Meanwhile, influential retweets are more effective than company tweets in bringing new Weibo followers to the company, which indirectly increases viewing. The authors discuss recommendations on how to manage tweeting as a marketing tool.
93 citations
TL;DR: The visual analytics pipeline for the social media is summarized, combining the above categories and supporting complex tasks and with these techniques, social media analytics can apply to multiple disciplines.
Abstract: With the development of social media e.g. Twitter, Flickr, Foursquare, Sina Weibo, etc., a large number of people are now using them and post microblogs, messages and multi-media information. The e...
85 citations
TL;DR: FOS, a fuzzy- based oversampling method that generates synthetic data samples from limited observed samples based on the idea of fuzzy-based information decomposition, is proposed and an ensemble learning approach that learns more accurate classifiers from imbalanced data in three steps is developed.
82 citations
TL;DR: A research model is developed to investigate the factors (affective cues in particular) that drive users to instantly share information on microblogs and explores the moderating role of gender and the results confirm the positive effects of informational, ambient, and social interactivity cues on individuals positive emotion, which subsequently promotes their urge to share information to microblogs.
Abstract: An impulsive-based model is proposed to explain instant information sharing.Information uniqueness and social interactivity increase positive emotion and urge.Males are stimulated by information uniqueness and information crowding.Social interactivity plays a dominant role in sparking the urge for female users.Males experience more positive emotion and engage in impulsive information sharing. Instant information sharing on microblogs is important for promoting social awareness, influencing customer attitudes, and providing political and economic benefits. However, research on the antecedents and mechanisms of such instant information sharing is limited. To address that issue, this study develops a research model to investigate the factors (affective cues in particular) that drive users to instantly share information on microblogs and explores the moderating role of gender. An online survey was conducted on a microblogging platform to collect data for testing the proposed research model and hypotheses. The results confirm the positive effects of informational (i.e., information uniqueness), ambient (i.e., information crowding), and social (i.e., social interactivity) cues on individuals positive emotion, which subsequently promotes their urge to share information on microblogs. Moreover, the moderating effects of gender are identified. This study contributes to the understanding of instant information sharing from an impulsive behavior perspective. The results also provide important insights for service providers and practitioners who wish to promote instant information sharing on microblogs.
TL;DR: A multimodal joint sentiment topic model (MJST) is proposed for weakly supervised sentiment analysis in microblogging, which applies latent Dirichlet allocation (LDA) to simultaneously analyze sentiment and topic hidden in messages based the introduction of emoticons and microbloggers personality.
31 Jul 2017
TL;DR: A novel machine learning based approach for automatic identification of the users spreading rumorous information by leveraging the concept of believability, i.e., the extent to which the propagated information is likely to be perceived as truthful, based on the trust measures of users in Twitter's retweet network is proposed.
Abstract: Ubiquitous use of social media such as microblogging platforms brings about ample opportunities for the false information to diffuse online. It is very important not just to determine the veracity of information but also the authenticity of the users who spread the information, especially in time-critical situations like real-world emergencies, where urgent measures have to be taken for stopping the spread of fake information. In this work, we propose a novel machine learning based approach for automatic identification of the users spreading rumorous information by leveraging the concept of believability, i.e., the extent to which the propagated information is likely to be perceived as truthful, based on the trust measures of users in Twitter's retweet network. We hypothesize that the believability between two users is proportional to the trustingness of the retweeter and the trustworthiness of the tweeter, which are two complementary measures of user trust and can be inferred from retweeting behaviors using a variant of HITS algorithm. With the retweet network edge-weighted by believability scores, we use network representation learning to generate user embeddings, which are then leveraged to classify users into as rumor spreaders or not. Based on experiments on a very large real-world rumor dataset collected from Twitter, we demonstrate that our method can effectively identify rumor spreaders and outperform four strong baselines with large margin.
TL;DR: Evaluating the content of messaging regarding the HPV vaccine on the social media and microblogging site Twitter and describing the sentiment of those messages offers a novel perspective to explore the context of health communication surrounding certain controversial issues.
Abstract: Objectives
Given the degree of public mistrust and provider hesitation regarding the human papillomavirus (HPV) vaccine, it is important to explore how information regarding the vaccine is shared online via social media outlets. The purpose of this study was to evaluate the content of messaging regarding the HPV vaccine on the social media and microblogging site Twitter, and describe the sentiment of those messages.
Design and Sample
This study utilized a cross-sectional descriptive approach. Over a 2-week period, Twitter content was searched hourly using key terms “#HPV and #Gardasil,” which yielded 1,794 Twitter posts for analysis. Each post was then analyzed individually using an a priori coding strategy and directed content analysis.
Results
The majority of Twitter posts were written by lay consumers and were sharing commentary about a media source. However, when actual URLs were shared, the most common form of share was linking back to a blog post written by lay users. The vast majority of content was presented as polarizing (either as a positive or negative tweet), with 51% of the Tweets representing a positive viewpoint.
Conclusions
Using Twitter to understand public sentiment offers a novel perspective to explore the context of health communication surrounding certain controversial issues.
TL;DR: The results show that the sarcasm detection task benefits from the inclusion of features which capture authorial style of the microblog authors, namely - function words and part of speech n-grams.
26 Oct 2017
TL;DR: This is the first study that tests the performance of research journals on Twitter according to their handles, observing how the dissemination of content in this microblogging network influences the citation of their papers.
Abstract: The purpose of this paper is to analyze the relationship between dissemination of research papers on Twitter and its influence on research impact.,Four types of journal Twitter accounts (journal, owner, publisher and no Twitter account) were defined to observe differences in the number of tweets and citations. In total, 4,176 articles from 350 journals were extracted from Plum Analytics. This altmetric provider tracks the number of tweets and citations for each paper. Student’s t-test for two-paired samples was used to detect significant differences between each group of journals. Regression analysis was performed to detect which variables may influence the getting of tweets and citations.,The results show that journals with their own Twitter account obtain more tweets (46 percent) and citations (34 percent) than journals without a Twitter account. Followers is the variable that attracts more tweets (s=0.47) and citations (s=0.28) but the effect is small and the fit is not good for tweets (R2=0.46) and insignificant for citations (R2=0.18).,This is the first study that tests the performance of research journals on Twitter according to their handles, observing how the dissemination of content in this microblogging network influences the citation of their papers.
Posted Content•
TL;DR: This paper develops a cascaded ensemble learning classifier for identifying the posts having racist or radicalized intent on Tumblr and shows that the proposed approach is effective and the emotion tone, social tendencies, language cues and personality traits of a narrative are discriminatory features for identifies the racist intent behind a post.
Abstract: Research shows that many like-minded people use popular microblogging websites for posting hateful speech against various religions and race. Automatic identification of racist and hate promoting posts is required for building social media intelligence and security informatics based solutions. However, just keyword spotting based techniques cannot be used to accurately identify the intent of a post. In this paper, we address the challenge of the presence of ambiguity in such posts by identifying the intent of author. We conduct our study on Tumblr microblogging website and develop a cascaded ensemble learning classifier for identifying the posts having racist or radicalized intent. We train our model by identifying various semantic, sentiment and linguistic features from free-form text. Our experimental results shows that the proposed approach is effective and the emotion tone, social tendencies, language cues and personality traits of a narrative are discriminatory features for identifying the racist intent behind a post.
TL;DR: Wang et al. as discussed by the authors examined the content moderation strategies of Sina Weibo, China's largest microblogging platform, in regulating discussion of rumors following the 2015 Tianjin blasts.
Abstract: China's control of information online is often imposed by social media platforms in the name of “rumor management.” This article examines the content moderation strategies of Sina Weibo, China's largest microblogging platform, in regulating discussion of rumors following the 2015 Tianjin blasts. More than 100,000 Weibo posts were collected and categorized into three data sets: rumor discussion posts from the public, rumor-debunking posts from Weibo's official rumor rebuttal accounts, and posts removed from the system. Two content-moderation rumor strategies, namely rumor rebuttal and content removal, were identified. Clustering analysis and time series analysis was applied to test how these two strategies were used to filter posts of different topics and how they were associated with public discussion of rumor-related topics. Our findings suggest that the platform's response to rumor varied depending on the political sensitivity of the topic. Time-series analysis indicated that the implementation of both strategies was usually associated with a subsequent increase in general discussion about the rumor, suggesting that these strategies do not create a consistent chilling effect on public speech.
TL;DR: It is found that the support of top managers, the access and competence of IT personnel, and the regional economic and social environments are key determinants of the emergence of open innovation in the public sector in China.
TL;DR: Examining how people interact in a popular microblogging-based learning community by examining a one-hour synchronous chat event and analyzing its network structure, levels of participation, major topics generated and types of interaction suggested that the levels of Participation were largely uneven.
Abstract: Research on microblogging in education has suggested its potential to promote community building and collaborative learning, but little is known about the nature of interaction in such microblogging communities. More research is needed to understand how online learning communities can be designed in a way that supports effective learning. The purpose of the study is to explore how people interact in a popular microblogging-based learning community by examining a one-hour synchronous chat event and analyzing its network structure, levels of participation, major topics generated and types of interaction. The findings suggested that the levels of participation in the synchronous chat were largely uneven. During the chat, participants were engaged in many types of interaction and discussed a variety of issues related to the topic. Interestingly, people who were different in their levels of participation also varied on the types of topics generated, but not on the types of interaction. [ABSTRACT FROM AUTHOR]
TL;DR: The results show that searching for relevant information on software applications within the vast stream of tweets can be compared to looking for a needle in a haystack, but this relevant information can provide valuable input for software companies and support the continuous evolution of the applications discussed in these tweets.
Abstract: Users of the Twitter microblogging platform share a considerable amount of information through short messages on a daily basis. Some of these so-called tweets discuss issues related to software and...
TL;DR: A joint framework to combine tag correlation and user social relation for microblog recommendation is investigated and an iterative updating scheme is developed to get the final tag-user matrix for computing the similarities between microblogs and users.
TL;DR: This paper proposes an innovative private trajectories release model and associated algorithms with differential privacy guarantees that considers both data privacy and data utility.
01 Jan 2017
TL;DR: In this article, the concept of effective word score (EFWS) was introduced to speed up the computation process for sentiment analysis on Twitter, and the EFWS heuristic was used to select the right training samples.
Abstract: As microblogging services like Twitter are becoming more and more influential in today's globalized world, its facets like sentiment analysis are being extensively studied. We are no longer constrained by our own opinion. Others' opinions and sentiments play a huge role in shaping our perspective. In this paper, we build on previous works on Twitter sentiment analysis using Distant Supervision. The existing approach requires huge computation resource for analyzing large number of tweets. In this paper, we propose techniques to speed up the computation process for sentiment analysis. We use tweet subjectivity to select the right training samples. We also introduce the concept of EFWS (Effective Word Score) of a tweet that is derived from polarity scores of frequently used words, which is an additional heuristic that can be used to speed up the sentiment classification with standard machine learning algorithms. We performed our experiments using 1.6 million tweets. Experimental evaluations show that our proposed technique is more efficient and has higher accuracy compared to previously proposed methods. We achieve overall accuracies of around 80% (EFWS heuristic gives an accuracy around 85%) on a training dataset of 100K tweets, which is half the size of the dataset used for the baseline model. The accuracy of our proposed model is 2–3% higher than the baseline model, and the model effectively trains at twice the speed of the baseline model.
01 Sep 2017
TL;DR: It has been shown that the proposed method significantly outperformed the Tumasjan's method, a well-recognized method for election prediction based on Twitter data, in predicting the result of the 2017 French presidential election.
Abstract: Twitter is a social network that lets users post their opinions about current affairs, share their social events, and interact with others. Twitter has now become one of the largest sources of news, with over 200 million active users monthly. This paper proposes a method to predict election results based on Twitter data analysis. The method extracts and analyses sentimental information from microblogs to predict the popularity of candidates. The proposed method was used for predicting the result of the 2017 French presidential election. It has been shown that the proposed method significantly outperformed the Tumasjan's method, a well-recognized method for election prediction based on Twitter data.
TL;DR: A novel method to improve topics learned from Twitter content without modifying the basic machinery of LDA is investigated, based on a pooling process which combines Information retrieval (IR) approach and LDA.
01 Oct 2017
TL;DR: This study aimed to perform sentiment analysis on Turkish and English Twitter messages using Doc2Vec using the Semi-Supervised learning method and the results were recorded.
Abstract: Twitter is one of the most popular microblog sites developed in recent years. Feelings are analysed on the messages shared on Twitter so that users ideas on the products and companies can be determined. Sentiment analysis helps companies to improve their products and services based on the feedback obtained from the users through Twitter. In this study, it was aimed to perform sentiment analysis on Turkish and English Twitter messages using Doc2Vec. The Doc2Vec algorithm was run on Positive, Negative and Neutral tagged data using the Semi-Supervised learning method and the results were recorded.
TL;DR: Twitter is used in this research to obtain public opinions regarding a bus rapid transit system (BRT) in Cali, Colombia and a traditional qualitative research design is used featuring a two-step process.
Abstract: Capturing public perceptions regarding transit systems is an essential part of creating a just and equitable service. Gathering such perceptions also holds the potential for identifying ways of increasing ridership and for identifying sources of transport-related social exclusion. Traditionally, these perceptions have been obtained from public participatory forums or via surveys on users’ opinions. However, increasingly, the use of alternative methods of obtaining public perceptions has included media and social media sources. Twitter is one such social media source and it is used in this research to obtain public opinions regarding a bus rapid transit system (BRT). Rather than relying upon purely automated data mining techniques for analyzing Twitter feeds, a traditional qualitative research design is used featuring a two-step process. First, a text mining procedure is used on the Twitter feeds filtered out by keywords relevant to the transit system, and secondly, a structured content analysis is applied to the entries. This method is applied to a case study on a Bus Rapid Transit system in Cali, Colombia. The results highlight concerns with safety, problems with the system’s infrastructure, and behavioral issues on the bus as primary points of discussion.
[...]
TL;DR: T-Hoarder is described: a framework that enables tweet crawling, data filtering, and which is also able to display summarized and analytical information about the Twitter activity with respect to a certain topic or event in a web-page.
TL;DR: In this article, the authors examined how students participated in these live chats, perceived benefits and challenges and how prior experience and preconceived perceptions of Twitter influenced the live chat experience and intentions for continued participation.
Abstract: This study presents two cases in which undergraduates were introduced to Twitter in their teacher preparation program as a means of developing a personal learning network. Twitter live chats are synchronous discussions that allow education stakeholders to discuss issues and share resources, engaging on potentially a global scale via the social networking platform. This study examines how students participated in these live chats, perceived benefits and challenges and how prior experience and preconceived perceptions of Twitter influenced the live chat experience and intentions for continued participation. Pre-activity reflections, student tweets and post-activity reflections were analyzed both qualitatively and quantitatively. While familiarity with Twitter varied, no participants had previously participated in a professional Twitter live chat; the majority of participants indicated a positive perception and intensions to continue participating in Twitter live chats. Plans for introducing, scaffolding and reflecting on initial Twitter live chat experiences are detailed and considerations and implications are discussed.