СИСТЕМА ПОРІВНЯЛЬНОГО АНАЛІЗУ ТЕКСТОВИХ ДОКУМЕНТІВ

Abstract

The article analyzes the work of existing intellectual systems of comparative analysis of electronic text documents and description of their basic features. The study proposes an intelligent system of comparative analysis of text documents using a new method for text matches based on construction of logic and linguistic models of text information. The system consists of five functional units. First unit is responsible for the text input into system. The second one is identification of components of logic and linguistic models of texts, which requires construction of the set of logic and linguistic models of all natural language sentences, which in its turn build every text. The aim of the third unit is to construct for each of two texts logic and linguistic model that has linguistic and semantic-syntactic components. Then consolidation and replacement of structural components of logic and linguistic models obtained on the previous stage of the system functioning are performed. The fifth unit is devoted to comparison of logic and linguistic models of text documents. As a result, we obtain the percentage match by style, by structure of complex syntactic units, by similarity of keywords and phrases, by topic and uniqueness.