Abstracts

Information retrieval systems of the Internet created on the basе of the thematic catalogue, give higher precise results. Despite of it, they have a few disadvantages, so for example, the creation and management of the thematic catalogue is executed by the person manually, and it requires huge intellectual activity and professionalism of staff, the updating of the database of such system is very laborious and long-continued process, with increase of quantity of subdirectories in the thematic catalogue complicates search process of the necessary information sources and obtaining of desired results. The retrieval systems based on automatic indexes, on the contrary, are more flexible and simple for implementation, cover more information sources, than thematic catalogues, i.e. the recall of automatic indexes much more exceeds recall of the thematic catalogues. However precision of search results of automatic index is lower than precision of the thematic catalogues.

The researches show, that the development of methods of creation of the automatic thematic catalogues without person's interference is actual and much promising. Here it is supposed, that the search robot of a system itself should determine subjects of information resource, the most eligible theme profiles from the thematic catalogues and to include resource into them. Thus the created thematic catalogues can be built according to present standards УДК, ББК etc. The created in this way automatic thematic catalogue can substitute the person (manager or operator of the retrieval system), will allow to execute activities on creation, service and updating of the thematic catalogue of the retrieval system.

In the present article the model of information search is researched. The methods of definition of information resources subjects, splitting of the information space on thematic directions and improvement of the thematic catalogue are considered. For the solution of formulated problems the theory of fuzzy sets and relations, and also approach Bellman-Zadeh are used.

Keywords: information retrieval, retrieval system, information resource, thematic catalogue, automatic index, automatic thematic catalogue, fuzzy sets, fuzzy relations.

Full Text in Russian:

PDF (301 kb)

Note. Abstracts are published in author's edition

Comments A.M.Fedotov

[Home] [Conferences] [SB RAS]

© 2001, Siberian Branch of Russian Academy of Science, Novosibirsk