Appendix 1: information management and evaluation.
The words that people use in communication are ill suited to standard word-based search engines. Frequently, new uses are invented for old words, ambiguous terms are used (the word "strike," for example, has more than 80 definitions), and multiple words are used to refer to the same idea. People also constantly categorize, but these categories are unstable both from individual to individual and from time to time, depending on the user's needs and interests. In short, human language is fuzzy and it requires fuzzy tools to deal with its meaning.
The problem is that most knowledge management tools are based not on how people use words, but on a symbolic approach to documents, where each word in a document is a symbol with a discrete and specific meaning. As a result, document retrieval systems that depend on the presence of exact words fail to retrieve relevant documents.
Biomimetic information management systems provide those fuzzy tools by using neural networks and other soft-computing techniques that emulate the way biological brains work. These systems also are self-organizing and do not require the laborious construction of rigid, expensive, prestructured rule bases. The result is an ad hoc categorization system that adapts itself to the intelligence problem at hand. The technology learns the meanings of words from the documents it indexes and can recognize the relevance of particular words or phrases based on their meaning. For example, if an investigator wants to search a document for the word "undercover," the system will not only indicate each place that the word "undercover" is used but also will find phrases such as "secret agent" or "covert operative." These systems are independent of the language in which the documents being searched are written.
Kintsch, W. & Keenan, J. (1973). Reading rate and retention as a function of the number of propositions in the base structure of sentences. Cognitive Psychology, 5, 257-274.
Barsalou, L. W. (1987) The instability of graded structure: Implications for the nature of concepts. In U. Neisser (Editor), Concepts and conceptual development, (pp. 101-140), Cambridge: Cambridge University Press.
Roitblat, H. L. & Henning, M. J. (1992) Connectionist investigations in language. International Congress of Psychology, Brussels, July, 1992, and Roitblat, H. L. (2001) Biomimetic Systems for Information Retrieval. In M. E. Williams (Ed.) Proceedings of the 22nd National Online Meeting (pp. 423-430). Medford NJ: Information Today.
Herbert L. Roitblat, Ph.D., at DolphinSearch, Inc. (firstname.lastname@example.org).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||biomimetic information management|
|Publication:||Countering Terrorism: Integration of Practice and Theory|
|Date:||Feb 28, 2002|
|Previous Article:||Conference participants.|
|Next Article:||Appendix 2: communications technologies.|