Argumentative Dialogue System for Product Selection in the Sales Industry

The paper presents the principles for the design of an intelligent assistant for the sales domain, which is intended to help the customer to find the right product. The user of the system is represented by a three-level model (general knowledge, intentions and representation on the current situation), supplemented by a number of objective and subjective characteristics of the user, formalised in the form of partial models. During the reasoning dialogue, the intelligent assistant identifies the user’s needs based on the user’s current intentions. Based on the user’s needs, the class of tasks to be solved by the user is determined. As a result, the intelligent assistant selects the required solution based on the user’s wishes and objections. The results of previous dialogues between the system and other users are used. The user’s reactions to the intelligent assistant’s decisions can correct the final result of the system’s work. The selection of the desired solution is made by comparing partial models formalising the objects of the domain, the classes of tasks and the user’s needs. The argumentation and justification of the found solution is based on the correspondence of the partial model of the proposed device with the formal models of the user’s needs and the classes of tasks to be solved. The intelligent assistant is able both to conduct a dialogue with the customer independently and to act as an advisor to the sales manager, prompting him with response options and providing references to previous solutions, depending on the context of the dialogue.