Automated methods for conducting an argumented dialogue with the user
Dmitry E. Palchunov, Alexander Yakobson · 2022 IEEE International Multi-Conference on Engineering, Computer and Information Sciences (SIBIRCON) · 2022
Human-computer interaction in the modern world has become a necessity, but the lack of user confidence in automatic systems creates great difficulties in their implementation. The problem is that in each case it is often impossible to understand the reasons for the decision. Therefore, the user has to try to find the arguments for this solution proposed by the system. The purpose of the work is to create an intelligent assistant capable of arguing his decisions, thereby strengthening the user's confidence in the correctness of his advice. The development of an intelligent assistant is based on a three-level semantic user model and an ontological domain model. The developed software package is designed to work as a virtual digital assistant in various fields. The implemented methods are based on the application of partial models and ontological homomorphisms. The system allows the user to search for the desired solution using the theory of argumentation and case-based reasoning, receiving a reasoned answer with a clear and transparent argumentation tree.