Texts Segmentation and Semantic Comparison: Method and Results of its Application

Abstract

The paper describes the approach of scientific texts comparing. The approach is characterized by the comparison of significant textual passages that consist of elements of meaning. The proposed method of scientific texts comparison differs from others known methods by the using of segmentation with semantic criteria, taking into account synonyms. That allows to automatically detect the semantic similarity between two compared texts, and take into account both the morphological structure of the text and its lexico-semantic content. The results of practical application of the method are also presented in the paper.

Keywords

• natural language processing
• plagiarism detection
• texts segmentation
• texts semantic comparison