Internet search and lexical-semantic processing of analogs when making decisions in various subject areas

• Main Author: S. F. Lipnitsky
• Format: Article
• Language: Russian
• Published: The United Institute of Informatics Problems of the National Academy of Sciences of Belarus 2021-01-01
• Series: Informatika
• Subjects: internet search; linguistic dictionaries; mathematical model; problem situation; relevance; query synthesis; tonal coloration
• Online Access: https://inf.grid.by/jour/article/view/1109

A mathematical model of Internet search and lexical-semantic processing of analogs of the decisions made in various subject areas is proposed. The search is carried out on queries synthesized from descriptions of problem situations. The synthesis process is carried out in two stages. At the first stage, sentences from the text containing a description of the problem situation are classified taking into account the informativeness of the verbal association between them. At the second stage, the information content of each class is calculated. The most informative of the generated classes, after their indexing, are used as queries. In lexical-semantic processing, the found analogs of solutions are examined for sentiment. When assessing sentiment, linguistic dictionaries of tone-colored vocabulary are used, which are formed on the basis of special tone-colored thematic corpora of texts. In the extreme case, two types of dictionaries are created. The first type is intended for the analysis of the positive sentiment in the descriptions of the decisions made, and the second is intended for the analysis of the negative sentiment, that is, in the process of lexical-semantic analysis, mainly positive and negative aspects of the decisions made are considered. The results of the analysis are presented to the user (decision-maker).