Abstract

The publicly available data such as the massive and dynamically updated news and social media data streams (a.k.a. big data) covers the various aspects of social activities, personal views and expressions, which points to the importance of understanding and discovering the knowledge patterns underlying the big data, and the need of developing methodologies and techniques to discover real-world events from such big data, to manage and to analyze the discovered events in an efficient and elegant way. In this talk we introduce techniques of discovering events from the multi-modal big data and building an event cube model to support event queries and analysis, by addressing the tasks of data cleansing, data fusion, event detection and modeling. Preliminary experimental results on some of the tasks will be reported. We further explore and connect the important events discovered in a multimodal collection of inputs from various public sources, uncover their co-occurrence and track down the spatial and temporal dependency to answer the challenging questions of "how" and "why". A novel event cube (EC) model is devised to support various queries and analysis tasks of events; such events include those discovered by techniques of untargeted event detection (UED) and targeted event detection (TED) from multi-sourced data. More specifically, based on essential event elements of 5W1H, the EC model is developed to organize the discovered events from multiple dimensions, and to operate on the events at various levels of granularity, which facilitates analyzing and mining hidden/inherent relationships among the events effectively.