Integrating Social and Auxiliary Semantics for Multifaceted Topic Modeling in Twitter

Abstract

Microblogging platforms, such as Twitter, have already played an important role in recent cultural, social and political events. Discovering latent topics from social streams is therefore important for many downstream applications, such as clustering, classification or recommendation. However, traditional topic models that rely on the bag-of-words assumption are insufficient to uncover the rich semantics and temporal aspects of topics in Twitter. In particular, microblog content is often influenced by external information sources, such as Web documents linked from Twitter posts, and often focuses on specific entities, such as people or organizations. These external sources provide useful semantics to understand microblogs and we generally refer to these semantics as auxiliary semantics. In this article, we address the mentioned issues and propose a unified framework for Multifaceted Topic Modeling from Twitter streams. We first extract social semantics from Twitter by modeling the social chatter associated with hashtags. We further extract terms and named entities from linked Web documents to serve as auxiliary semantics during topic modeling. The Multifaceted Topic Model (MfTM) is then proposed to jointly model latent semantics among the social terms from Twitter, auxiliary terms from the linked Web documents and named entities. Moreover, we capture the temporal characteristics of each topic. An efficient online inference method for MfTM is developed, which enables our model to be applied to large-scale and streaming data. Our experimental evaluation shows the effectiveness and efficiency of our model compared with state-of-the-art baselines. We evaluate each aspect of our framework and show its utility in the context of tweet clustering.

A framework for event classification in tweets based on hybrid semantic enrichment

“…semantics, cooccurrences, representative terms) are retrieved and combined with the textual features. Criteria used to select textual tokens to be enriched are named entities (SAIF;HE;ALANI, 2012;ABEL et al, 2012a;GUCKELSBERGER;JANSSEN, 2015;VOSECKY et al, 2014), frequent or relevant terms measured using Term Frequency -Inverse Document Frequency (TF-IDF) (PACKER et al, 2012;PAULHEIM, 2013;VOSECKY et al, 2014), and location/time identification heuristics GUCKELSBERGER;JANSSEN, 2015;VOSECKY et al, 2014). For the contextual enrichment, different techniques and elements can be explored:…”

“…• External documents: context is provided by the content extracted from related web docu-ments, which can be identified through Uniform Resource Locators (URL) mentioned in the tweets (VOSECKY et al, 2014), or by selecting specific words in the messages (e.g. named entities, representative terms) to access, for example, related Wikipedia pages (GENC;SAKAMOTO;NICKERSON, 2011;ROSA et al, 2011).…”

“…Open Calais, Alchemy, DBpedia Spotlight) are employed to identify entities that are mentioned in the tweet text and provide related categorical information (SAIF;HE;ALANI, 2012;VOSECKY et al, 2014). Some NER tools are related to proprietary knowledge sources (e.g.…”

“…The content of tweet text is very informal and several tricks are used by Twitter users to convey meaning to their posts, such as hashtags, emoticons, and abbreviation. To identify the important pieces that compose a Twitter message, different approaches can be employed (SAIF; HE; ALANI, 2012; MOHAMMAD; KIRITCHENKO; ZHU, 2013; SCHULZ; RIS-TOSKI; PAULHEIM, 2013;ABEL et al, 2012a;GUCKELSBERGER;JANSSEN, 2015;VOSECKY et al, 2014;PACKER et al, 2012), among them:…”

Hybrid Topic Cluster Models for Social Healthcare Data

“…Twitter, Facebook, and microblogs [21], [22], [23] reveals the opinions of public and assessment of this social data [13], [14] is an emerging need in the applications like topics detection [1], [4], product promotion in business [4], political predictions [6], and health recommendations [2], [6]. Rapid urbanization is posing the number of public health-related problems, including accidents and injuries, healthcare disparities, and increasing disease burdens due to changes in lifestyle and nutrition, as well as increased environmental pollution.…”