Sentiment Analysis of Information Space as Feedback of Target Audience for Regional E-Business Support in Ukraine
Autor(en): | Vysotska, Victoria Markiv, Oksana Tchynetskyi, Stepan Polishchuk, Bohdan Bratasyuk, Oksana Panasyuk, Valentyna |
Herausgeber: | Emmerich, M. Leiden University Leiden Institute of Advanced Computer Science Niels Bohrweg 1 Leiden Vysotska, V. Osnabruck University Friedrich-Janssen-Str. 1 Osnabruck Lytvynenko, V. Kherson National Technical University Beryslavske Shosse 24 Kherson |
Stichwörter: | comment; Competition; content analysis; Data privacy; E- commerces; e-business; e-commerce; E-learning; eBusiness; Electronic commerce; feedback; information security; Logistic regression; Logistic regression method; machine learning; Machine-learning; NLP; personal data; personal data protection; Personal data protections; Sentiment analysis; Support vector machines; Target audience; Ukraine | Erscheinungsdatum: | 2023 | Herausgeber: | CEUR-WS | Enthalten in: | CEUR Workshop Proceedings | Band: | 3426 | Startseite: | 488 – 513 | Zusammenfassung: | In conditions of the war in Ukraine, e-business plays a key role in supporting and developing the economy of country, maintaining business relations and competitiveness in the international area of the financial market, interacting with government bodies and supporting feedback from the target audience. The paper describes the application of sentiment analysis of comments, feedback, requests and news for the support and development of e-business. The analyzed analogs made it possible to develop information technology for solving NLP problems of e-business, adapted for the Ukrainian target audience. The general typical structure of the information system for the support and development of e-commerce has been developed by analyzing the feedback of the target audience based on machine learning technology and natural language processing methods. The logistic regression method coped best with the task of analyzing the impact of the news on the financial market, which has shown an accuracy of 75.67%. This is certainly not the desired result, but it is the largest indicator of all the considered. The support vector method (SVM) has shown an accuracy of 72.78%, which is a slightly worse result than the one obtained with the help of the logistic regression method. And the naive Bayesian classifier method has shown the worst accuracy of 71.13%, which is less than the two previous methods. © 2023 Copyright for this paper by its authors. |
Beschreibung: | Cited by: 0; Conference name: 2023 Modern Machine Learning Technologies and Data Science Workshop, MoMLeT and DS 2023; Conference date: 3 June 2023; Conference code: 189914 |
ISSN: | 1613-0073 | Externe URL: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85164906943&partnerID=40&md5=761fa955816210b2a3f3efee1978f436 |
Zur Langanzeige