User and Topic Analytics of the Social Web of News
Abstract
The proliferation of social media is undoubtedly changing the way people produce and consume news online. Editors and publishers in newsrooms need to understand user engagement and audience sentiment evolution on various news topics. News consumers want to explore public reaction on articles relevant to a topic and refine their exploration via related entities, topics, articles and tweets. I will present SPAUE and MAQSA, two systems for social analytics on news. SPAUE analyzes traffic while MAQSA is content-centric. The core data model in SPAUE is a graph formed by the collective behavior of users represented as a set of actions such as browsing, posting an opinion and sharing news stories. SPAUE implements path traversal primitives that aggregate topics and actions for a given time period and along four dimensions: traffic source, visits, visitors, and geographic location. MAQSA provides an interactive topic-centric dashboard that summarizes social activity around news articles. The dashboard contains an annotated comment timeline, a social graph of comments, and maps of comment sentiment and topics. Both SPAUE and MAQSA rely on scalable algorithms that enable an interactive specification of topics, actions, and dates and dynamically process large collections of relevant articles.