A Russian-Language Text Summarization System

With the development of information technologies, time costs necessary to find information of interest increase, which reduces the personal effectiveness of the user. In such conditions, the field of automatic summarization becomes more and more relevant, since this sphere of natural language processing allows you to provide the necessary information contained in sources in a reduced form, which saves time resources. In this research, a program was prepared that allows you to generate a summary from the arrays of news articles. The result of automatic summarization is a document that contains the relevant information repeatedly appearing in the collection, without repeats, which at the same time includes additional information specific to any of the texts in the array. For the correct definition of the informative content of the text, a corpus of informative words for sports news was created. The collection was carried out with the help of crawlers, which save the text of the article and all the necessary metadata in a database. To analyze the work of the algorithm, articles from "RIA Novosti", "Interfax", "Korrespondent" and others were cited. Evaluation of each summary provided that it could be used as a tool for determining the quality of its work.