Issue No. 04 - July/August (1999 vol. 14)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/5254.784087
Many important information-processing applications involve text-based messages that multiple human observers generate over time. The tracking problem is to associate messages about the same object or phenomenon and create the most probable tracks based on those associations. This article describes a working prototype, TextTrack, for solving such problems. TextTrack combines mature radar signal-processing tracking methods with statistical models of natural-language messages.
Correlation, Natural Language, Bayesian, Fuzzy, Ambiguity, Tracking
George V. Cybenko, Carlos Alberola, "Tracking with Text-Based Messages", IEEE Intelligent Systems, vol. 14, no. , pp. 70-78, July/August 1999, doi:10.1109/5254.784087