Automatic Story Evolution Wikification from Social Data

Abstract

We present the generation of a new and dynamic data asset that captures the evolution of a story from different perspectives. In contrast to news articles that are ranked by relevance and freshness in a search engine or a static Wikipedia article that provides an overview of the event or topic, our solution consists of the automatic construction of a wiki-like document that highlights the salient items of a topic as it evolves over time, with related pivots that allow the user to explore related stories. We demonstrate the effectiveness of our approach by processing a dataset comprising millions of English language tweets generated over a one year period.