Project Chorologos

  • Formulation of expressive query types that enable selection of underlying spatio-temporal-textual data based on diverse information needs, going beyond exact or syntactical matching and towards semantic retrieval. Examples of such queries include similarity matching, pattern-based matching, as well as semantic similarity matching.
  • Theoretical contributions in terms of properties and search bounds for the proposed query types, thus laying the foundations for efficient processing and search.
  • Design of appropriate access methods that jointly index space, time, and text, in an appropriate way to support filtering of data that is irrelevant to the query at hand.
  • Efficient query processing algorithms following well-established methodologies, including filter-and-refine and branch-and-bound, aiming at fast delivery of accurate query results.
  • Parallel processing of the proposed query types, towards scalable algorithms that make the analysis of vast-sized data sets feasible in practice.

Read more