Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Generalized lsa method and its use for assessing the significance of fuzzy collocation in text collections D.V. POLYAKOV, N.M. MITROFANOV, E.N. LEPJOSHKIN The article formulated the general statement of a problem of formalizing of text’s collections for its subsequent use in the algorithms of information retrieval and clustering text collections. Adequacy of this generalization is proved by considering the frequent cases of formalizing the collection, that are well-known and well-researched approaches. A generalized vector-space model of textual collections and investigated the possibility of using the method of latent semantic analysis in this model. A term factor that reflects the semantic component of a text document is given. And method for assessing of the semantics’ significance of the factor of the text document is offered. Also, in the article formulated an approach to the formulation and implementation of computational experiments to carry out a comparative assessment of the semantic significance of various factors. The article has not address issues related to the application of the developed models and methods for solving the problems of information retrieval and clustering text collections. Keywords: information retrieval, clustering, text collection, latent semantic analysis, factor analysis, formalization of a text, the theory of fuzzy sets. Contacts: E-mail: dimadress@yandex.ru Pp. 45-55.