Task 3 Web Community Sensing & Task 6 Query and Visualization REACTION Workshop January 31 th, 2013
Summary of on-going activities Team update WP3 & WP6 progress reports Resources & publications
Team update Former collaborators: Nuno Baldaia PhD @ FEUP UT-Austin, REACTION grant Current collaborators: Arian Pasquali Researcher @ SAPO Labs U. Porto, MSc @ DCC-FCUP Gustavo Laboreiro PhD @ LIACC, REACTION grant Jorge Teixeira - Researcher @ SAPO Labs U. Porto, PhD @ LIACC Luís Rei - Researcher @ SAPO Labs U. Porto Eugénio Oliveira Prof. Cat. @ FEUP & LIACC Eduarda Mendes Rodrigues Prof. Aux. @ FEUP (on leave from March 13) New team members: Jorge Moreira 4 month REACTION grant (started Dec 12), MSc @ DEI-FEUP Tiago Cunha 8 month REACTION grant (started Jan 19), MSc @ DEI-FEUP Carlos Soares Prof. Ass. @ FEUP & INESC TEC (March 13 onwards)
WP3 progress report 1. Modeling the credibility and authority of news sources and opinion makers in social networks Content pre-processing: text tokenization, error correction, tokenization, bot detection algorithms language variant detection for community filtering social network features to be explored enrich micro-blog texts with external information context from news (e.g. topics, events) User profiling: network analysis metrics and visualizations (e.g., user centrality in the #codebits reply-to network) mining interaction patterns in social media content for automatic user classification mining spatial-temporal patterns
WP3 progress report 2. Identifying influential individuals and experts on a given news topic Experiments with topic-based influence detection algorithms to be integrated in TwitterEcho TwitterEcho analytics modules exploratory analysis of user activity and topic profiles visual analytics dashboard crawl monitor, community analysis, geo-located infovis, time series, etc. search interfaces for topic-focused analysis and retrieval
WP3 progress report 3. Monitoring the community reaction to news stories and the polarity of opinions POPmine platform message retrieval, opinion mining and aggregation of social media & online news (POPSTAR project) cross-media analysis (news, tweets, blogs) for any personality from SAPO Verbetes to be used in conjunction with TwitterEcho SentiBubbles visualization of public opinion in social media (Facebook, Twitter) based on the SentiLex-PT corpus
WP6 progress report 1. Development of tools for querying extracted information and visualizing annotated documents and datasets Entity extraction with SAPO Verbetes Citation extraction with SAPO Voxx Extraction and visualization of news networks (MVDi, Mundo numa Rede) and news profiles for Verbetes entities TwitterEcho dashboard 2. Continuous scanning of the social web, news sources and various kinds of data streams Online news media and news archives TwitterEcho crawler modular architecture for covering real-time stream crawling of any community, topic or event SAPO Blogs, Nutch crawler for Web forums and blogs
Resources & publications Computational Newsroom tools: TwitterEcho 3 (+ POPmine) crawler, data mining, dashboard MVDi ( Mundo Visto Daqui interactivo) widgets, O Mundo numa Rede SentiBubbles, Twitteuro, Twitómetro widgets NetViz of topic-based Twitter influencers Publications: G. Laboreiro, M. Bošnjak, E. Mendes Rodrigues, L. Sarmento and E. Oliveira. Determining language variant in microblog messages. Proc. The 28th ACM Symposium On Applied Computing, SAC 2013, Information Access and Retrieval Track (IAR). Bosnjak, M., Sarmento, L., and Mendes Rodrigues, E. Robust Language Identification with RapidMiner - A Text Mining Use Case. To appear in: Hofmann, M. and Klinkenberg, R. (Eds.), Use Cases with RapidMiner (final revision).
Technical presentations Talk: Verbetes v3 (Luís Rei) Short presentations: News media analysis News information extraction and visualization (Jorge Teixeira) Social media analysis TwitterEcho 3 - twitter research platform (Arian Pasquali) Content pre-processing (Gustavo Laboreiro) Spatial-temporal data mining (Tiago Cunha) Information extraction from social media (Jorge Moreira)