P. P. Kokorin

At the present time, library automation systems do not support technologies of automated semantic text analysis as well as modern text analysis systems are not adapted to work with library automation systems. This makes desirable to expand facilities of library automation systems by means of a semantic text analysis ability. The article outlines the principles of construction of an infological system for analytical monitoring of the scientific and technical literature in electronic libraries. The distinction of using an electronic library instead of other thematic collections of texts as a source for the construction of subject areas anthologies is shown. Anthology is an essential element for building infological system, and its the quality determines the efficiency of the entire system as a whole. A systematic process of forming the library fund provides full, accurate, and authoritative anthology of subject areas, which in turn allows infological systems to create hierarchical structure of semantically related notions (thesaurus of a subject area). Such thesauri can be used to monitor materials in the library, analytical texts annotating, and gives a number of new search features for electronic materials (associative search, search for similar ones, duplicates, etc.). In addition, those anthologies based on scientific and technical texts of electronic library allow one to achieve a higher level of accuracy and reliability of search results for Internet search engines. The proposed solution based on an assessment of similarity of thesauri of a document and subject area enables to assess the adequacy, relevance and reliability of a document to a subject area topics. This measure of similarity is taken into account in creating a search index, which ultimately affects the relevancy of search results. The experiments have shown the high efficiency of the method of analytical technical texts annotating. Also for the task of monitoring scientific and technical materials of an Internet application the thesauri of subject areas have shown high efficiency in assessing the adequacy, relevance and reliability of documents are being added to a search engine index, thus improving the quality of search engine indexing of documents and improve the relevancy of search results