Knowledge Triggering, Extraction and Storage via Human-Robot Verbal Interaction

Abstract

This article describes a novel approach to expand in run-time the knowledge base of an Artificial Conversational Agent. A technique for automatic knowledge extraction from the user's sentence and four methods to insert the new acquired concepts in the knowledge base have been developed and integrated into a system that has already been tested for knowledge-based conversation between a social humanoid robot and residents of care homes. The run-time addition of new knowledge allows overcoming some limitations that affect most robots and chatbots: the incapability of engaging the user for a long time due to the restricted number of conversation topics. The insertion in the knowledge base of new concepts recognized in the user's sentence is expected to result in a wider range of topics that can be covered during an interaction, making the conversation less repetitive. Two experiments are presented to assess the performance of the knowledge extraction technique, and the efficiency of the developed insertion methods when adding several concepts in the Ontology.

Knowledge-Grounded Dialogue Flow Management for Social Robots and Conversational Agents

“…The approach presented in this article has obvious limitations since it relies on an Ontology of concepts and the related sentences to talk about such concepts, which needs to be manually encoded by experts. This issue has been addressed in subsequent work, which explored strategies for expanding the knowledge base at run-time during the interaction with the user [52].…”

“…To exploit the information related to the category of the sentence, the Content Classification of the topics contained in the Ontology is a fundamental operation and needs to be performed in a setup phase to create a mapping between the topics contained in the Ontology and the CNL category hierarchy. Off-line, before the system starts, the classification procedure is performed for all the topics: the algorithm puts together all the sentences associated with each topic (inserted into the Ontology by experts and then automatically composed, or added by the users during the conversation through a mechanism not described in this article [52]), sends them to CNL, and associates the returned categories to the corresponding topic in the Ontology and then in the DT (notice that, whilst keywords are manually encoded in the Ontology, categories are automatically assigned). On-line, during the conversation, this mapping will allow the system to find which topics of the Ontology match best (i.e., have more categories in common) with the category (if any) of the sentence pronounced by the user according to CNL.…”