This collaborative project "TANGO" is leveraging strengths of research teams at Brigham Young University (IIS-0414644, PI: David Embley) and Rensselaer Polytechnic Institute (IIS 0414854, PI: George Nagy). TANGO is a framework for organizing domain-specific factual data appearing in independently generated web pages. Algorithms and software are developed for extracting and interpreting individual lists and tables and integrating them with the contents of other tables that present partially overlapping information. The input pages may be HTML, PDF, PostScript, or scanned document image files. The output of the system is an ontology that provides the conceptual framework for the domain. The ontology evolves as web pages with additional relevant data are harvested and processed. Constructing ontologies is currently a labor and skill intensive process. The TANGO framework automates much of the ontology construction task for a domain when information exists in lists and tables that describe the domain. The results of this project build stepping stones to the Semantic Web, for which ontologies are a central component. Therefore, this project will have an impact on users of the Semantic Web, who assisted by the automated software agents developed in this project, can accomplish information-based tasks for research, scholarship, planning, design, education, and entertainment. This cross-disciplinary and cross-university endeavor introduces graduate and undergraduate students to cutting-edge research. Developed tools (including source programs), a corpus of tables both in raw and normalized forms, collected test data, and technical reports can be found at the project website (www.tango.byu.edu).
