DOI: 10.17587/prin.10.38-46

Software System for Text Sentiment Analysis Based on Sentiment Lexicons

M. V. Pletneva, pletneva.mv.kirov@gmail.com, Vyatka State University, Kirov, 610000, Russian Federation

Corresponding author: Pletneva Maria V., Assistant, Vyatka State University, Kirov, 610000, Russian Federation, E-mail: pletneva.mv.kirov@gmail.com

The article is devoted to the problem of automatic text sentiment analysis. Under the text sentiment the authors emotional attitude, expressed in the text is meant. There is a wide range of areas where sentiment analysis is applied, including political, sociological and marketing research, the search engines, the human-computer interfaces. The task of sentiment analysis is complicated by the strong dependence of the sentiment expressions on the context and domain, the presence of sarcasm and irony. To solve this problem the following approaches are used: a knowledge-based approach; machine learning and combined approach. Each of these approaches has advantages and disadvantages. In this article a combined approach is used. The article proposes the structure of the software system for text sentiment analysis based on sentiment lexicons. The structure includes four main subsystems, namely text preprocessing, sentiment lexicon generation, sentiment lexicon parameters determination and classification subsystems. The implementation of this system is considered in accordance with the object-oriented approach. Main classes and their relations are described with UML class diagram. The results of developed systems experimental research are presented. The research is performed using the text corpora of movies and books reviews, provided by organizers of ROMIP-2011 and ROMIP-2012 seminars. These results show the superiority of the proposed system over widespread machine learning methods. The developed system can be used as a standalone application for summarizing opinions on the text corpora. It is possible to integrate the main part of the system into search services, focused on the preparation of recommendations based on the text sentiment analysis.

Keywords: natural language processing, opinion mining, text sentiment analysis, sentiment lexicons, software system, class diagram

pp. 38–46

For citation:
Pletneva M. V. Software System for Text Sentiment Analysis Based on Sentiment Lexicons, Programmnaya Ingeneria, 2019, vol. 10, nо. 1, pp. 38—46'

This work was supported by the Ministry of Education and Science of the Russian Federation (the project No. 34.2092.2017/4.6 "Research and Development of Sentiment Lexicons for Text Sentiment Analysis")