Bibliometric analysis of chatbot training research: key concepts and trends

- Ляшенко, Роман Олегович (orcid.org/0009-0000-2614-6997) and Семеріков, С.О. (orcid.org/0000-0003-0789-0272) (2024) Bibliometric analysis of chatbot training research: key concepts and trends Information Technologies and Learning Tools, 3 (101). pp. 181-199. ISSN 2076-8184

Text 5622.pdf - Published Version Download (1MB)

Abstract

This bibliometric analysis aims to identify current research directions and priorities in the field of chatbot training – software agents capable of natural language dialogue. The study is based on the analysis of 549 scientific sources from the Scopus database on this topic. The analysis revealed a steady increase in relevant publications starting from 2018, indicating a growing relevance of this subject area in recent years. Based on a cluster analysis of keywords, four main research areas were identified: natural language processing, application of relevant technologies in various spheres of society, use of machine learning methods for natural language processing, and application of chatbots in education and services. In the field of natural language processing, the focus of current research is on computational linguistics, language modeling and machine comprehension, particularly speech recognition tasks. Leading research on artificial intelligence applications in this area concerns the responsible and ethical use of modern large language models and conversational agents, such as ChatGPT, in education and healthcare. Machine learning methods are actively being developed for creating virtual intelligent assistants, natural language user interfaces, and other natural language processing systems, including for diagnostic tasks in medicine. Key applications of chatbots are identified in adaptive learning systems, knowledge management, and customer service. Based on the analysis, the most significant concepts in each of the studied areas are defined to outline priorities for further research in the field of chatbot training. Future work involves conducting a systematic literature review with the automation of certain stages using large language models. In particular, these models will be employed to automatically classify study abstracts according to inclusion/exclusion criteria during the screening phase. Automating systematic review stages using artificial intelligence opens up significant prospects for accelerating scientific research, particularly in the education field based on an evidence-based approach.

Item Type:	Article
Keywords:	chatbot training; natural language processing; machine learning; bibliometric analysis; systematic literature review; large language models
Subjects:	Science and knowledge. Organization. Computer science. Information. Documentation. Librarianship. Institutions. Publications > 00 Prolegomena. Fundamentals of knowledge and culture. Propaedeutics > 004 Computer science and technology. Computing. Data processing > 004.9 ІКТ ( Application-oriented computer-based techniques )
Divisions:	Institute for Digitalisation of Education > Department of Cloud-Oriented Systems and Artificial Intelligence in Education
Depositing User:	Сергій Олексійович Семеріков
Date Deposited:	21 Jul 2024 10:47
Last Modified:	21 Jul 2024 10:47
URI:	https://lib.iitta.gov.ua/id/eprint/741785

Downloads

Actions (login required)

View Item