An official website of the United States government
Textual information, such as news articles, social media, and online forum discussions, often comes in a form of sequential text streams. Events happening in the real world trigger a set of articles talking about them or related events over a period of time. In the meanwhile, even one event is fading out, another related event could raise public attention. Hence, it is important to leverage the information about how topics influence each other over time to obtain a better understanding and modeling of document streams. In this paper, we explicitly model mutual influence among topics over time, with the purpose to better understand how events emerge, fade and inherit. We propose a temporal point process model, referred to as Correlated Temporal Topic Model (CoTT), to capture the temporal dynamics in a latent topic space. Our model allows for efficient online inference, scaling to continuous time document streams. Extensive experiments on real-world data reveal the effectiveness of our model in recovering meaningful temporal dependency structure among topics and documents.
more »
« less
Spatio-Temporal Event Detection from Multiple Data Sources
The proliferation of Internet-enabled smartphones has ushered in an era where events are reported on social media websites such as Twitter and Facebook. However, the short text nature of social media posts, combined with a large volume of noise present in such datasets makes event detection challenging. This problem can be alleviated by using other sources of information, such as news articles, that employ a precise and factual vocabulary, and are more descriptive in nature. In this paper, we propose Spatio-Temporal Event Detection (STED), a probabilistic model to discover events, their associated topics, time of occurrence, and the geospatial distribution from multiple data sources, such as news and Twitter. The joint modeling of news and Twitter enables our model to distinguish events from other noisy topics present in Twitter data. Furthermore, the presence of geocoordinates and timestamps in tweets helps find the spatio-temporal distribution of the events. We evaluate our model on a large corpus of Twitter and news data, and our experimental results show that STED can effectively discover events, and outperforms state-of-the-art techniques.
more »
« less
- PAR ID:
- 10143407
- Date Published:
- Journal Name:
- Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Twitter bot detection is vital in combating misinformation and safeguarding the integrity of social media discourse. While malicious bots are becoming more and more sophisticated and personalized, standard bot detection approaches are still agnostic to social environments (henceforth, communities) the bots operate at. In this work, we introduce community-specific bot detection, estimating the percentage of bots given the context of a community. Our method{---}BotPercent{---}is an amalgamation of Twitter bot detection datasets and feature-, text-, and graph-based models, adjusted to a particular community on Twitter. We introduce an approach that performs confidence calibration across bot detection models, which addresses generalization issues in existing community-agnostic models targeting individual bots and leads to more accurate community-level bot estimations. Experiments demonstrate that BotPercent achieves state-of-the-art performance in community-level Twitter bot detection across both balanced and imbalanced class distribution settings, presenting a less biased estimator of Twitter bot populations within the communities we analyze. We then analyze bot rates in several Twitter groups, including users who engage with partisan news media, political communities in different countries, and more. Our results reveal that the presence of Twitter bots is not homogeneous, but exhibiting a spatial-temporal distribution with considerable heterogeneity that should be taken into account for content moderation and social media policy making.more » « less
-
With the spread of the SARS-CoV-2, enormous amounts of information about the pandemic are disseminated through social media platforms such as Twitter. Social media posts often leverage the trust readers have in prestigious news agencies and cite news articles as a way of gaining credibility. Nevertheless, it is not always the case that the cited article supports the claim made in the social media post. We present a cross-genre ad hoc pipeline to identify whether the information in a Twitter post (i.e., a “Tweet”) is indeed supported by the cited news article. Our approach is empirically based on a corpus of over 46.86 million Tweets and is divided into two tasks: (i) development of models to detect Tweets containing claim and worth to be fact-checked and (ii) verifying whether the claims made in a Tweet are supported by the newswire article it cites. Unlike previous studies that detect unsubstantiated information by post hoc analysis of the patterns of propagation, we seek to identify reliable support (or the lack of it) before the misinformation begins to spread. We discover that nearly half of the Tweets (43.4%) are not factual and hence not worth checking – a significant filter, given the sheer volume of social media posts on a platform such as Twitter. Moreover, we find that among the Tweets that contain a seemingly factual claim while citing a news article as supporting evidence, at least 1% are not actually supported by the cited news, and are hence misleading.more » « less
-
Social media is being increasingly utilized to spread breaking news and updates during disasters of all magnitudes. Unfortunately, due to the unmoderated nature of social media platforms such as Twitter, rumors and misinformation are able to propagate widely. Given this, a surfeit of research has studied rumor diffusion on social media, especially during natural disasters. In many studies, researchers manually code social media data to further analyze the patterns and diffusion dynamics of users and misinformation. This method requires many human hours, and is prone to significant incorrect classifications if the work is not checked over by another individual. In our studies, we fill the research gap by applying seven different machine learning algorithms to automatically classify misinformed Twitter data that is spread during disaster events. Due to the unbalanced nature of the data, three different balancing algorithms are also applied and compared. We collect and drive the classifiers with data from the Manchester Arena bombing (2017), Hurricane Harvey (2017), the Hawaiian incoming missile alert (2018), and the East Coast US tsunami alert (2018). Over 20,000 tweets are classified based on the veracity of their content as either true, false, or neutral, with overall accuracies exceeding 89%.more » « less
-
Twitter is an extremely popular micro-blogging social platform with millions of users, generating thousands of tweets per second. The huge amount of Twitter data inspire the researchers to explore the trending topics, event detection and event tracking which help to postulate the fine-grained details and situation awareness. Obtaining situational awareness of any event is crucial in various application domains such as natural calamities, man-made disaster and emergency responses. In this paper, we advocate that data analytics on Twitter feeds can help improve the planning and rescue operations and services as provided by the emergency personnel in the event of unusual circumstances. We take an emotional change detection approach and focus on the users’ emotions, concerns and feelings expressed in tweets during the emergency situations, and analyze those feelings and perceptions in the community involved during the events to provide appropriate feedback to emergency responders and local authorities. We employ improved emotion analysis and change point detection techniques to process, discover and infer the spatiotemporal sentiments of the users. We analyze the tweets from recent Las Vegas shooting (Oct. 2017) and note that the changes in the polarity of the sentiments and articulation of the emotional expressions, if captured successfully can be employed as an informative tool for providing feedback to EMS.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1007/978-3-030-16148-4_23
