Towards a Metamodel for Proactive Chatbots
Charaf Ouaddi, Lamya Benaddi, Adnane Souha, Abdeslam Jakimi
Abstract
Recently, increasing popularity has been gained by chatbots in various activity fields as they can automate customer service and reduce the effort of humans. However, the effectiveness of chatbots is still not satisfactory because some drawbacks are associated with these applications, such as NLP service lock-in and high costs linked to their implementation platforms. Therefore, one of the possible solutions to solve this issue in chatbot development is the development of a domain-specific language (DSL), which is defined through three main components: abstract syntax, concrete syntax, and semantics. In this paper, the proactive behavior within chatbots is pre-sented, and the proactive mechanisms offered by some chatbot development tools are discovered. Thus, a new metamodel for proactive assistance based on the Rasa framework is proposed. The proposed metamodel integrates all the components (concepts, classes, attributes, and relationships) that facilitate the building of DSL for chatbots.