About STempo

Updated on August 27, 2014 

A combination of visual and computational tools combined with human knowledge and intuition is needed to deal with the complexities of finding meaningful patterns - of separating the noise from true temporal and spatio-temporal associations and revealing the complex structure of dynamic associations hidden in the wealth of space-time data now available on the Web, such as news feeds and social media streams.

The goal of the STempo project is to develop and implement computational tools for revealing complex associations in reported events. The prototype being developed represents a tightly integrated temporal and spatio-temporal pattern discovery environment that synergistically combines the power and impartiality of statistical methods to uncover hidden and potentially unsuspected patterns in complex data with the capabilities of the human analyst to see patterns visually and, in turn, apply prior knowledge to guide analysis.

We have designed and implemented a computational / visualization framework that allows the analyst to work with both the data and computational/statistical tools in a highly interactive and iterative manner.  The figure below shows the latest main STempo view, with newly-revised visual tools that provide entity-based, location-based, time-based and event-based representations. These are tightly linked and are used for filtering and selecting the data for both closer visual inspection and for input to computational analysis. A statistical/computational technique for discovering hierarchical associations among events and groups of events, called T-pattern analysis, has been implemented with a number of extensions to accommodate the complexities of real-world event data.  We have also developed and extended a pattern matching process that provides a quantitative measure of pattern consistency or change in event data over differing locations and/or times, once a pattern has either been identified via T-pattern analysis or entered directly by the user.

Research Goals

  • new procedures that represent the generalization of the small set of truly inductive and scalable computational pattern discovery techniques
  • the integration of advanced visualization with innovative computational capabilities for inductively finding temporal and spatio-temporal patterns in data from diverse sources,
  • design principles for how to integrate and relate diverse visualizations for representing events in the temporal, spatial and object dimensions in a way that is useful and intuitive for the human analyst
  • a running STempo software prototype that demonstrates the viability and robustness of the conceptual advancements above