Year: 2024
Type: article
Abstract: In the context of the rapid digitalization of educational processes, automation of interaction with applicants is becoming a priority task for universities. This article is devoted to the analysis of various approaches to the development of chatbots for the admissions committee. Both traditional methods based on rules and keywords, and modern technologies using machine learning and large language models are considered. Particular attention is paid to comparing the accuracy, relevance and naturalness of responses, as well as implementation and support costs. The study shows that modern neural network models are significantly superior to classical approaches in a number of key parameters, but require large resources and constant Internet access. Combining classical and neural network methods can be the optimal solution, allowing you to effectively use the strengths of both approaches, while ensuring high quality and efficiency of interaction with applicants.
Authors: Aleksei Rodionov, Daniil Korzh
Institution: Baikal State University
Language: English
Topic: AI in Service Interactions
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: bronze
OpenAlex ID: W4411326907