Equipping a Conversational Agent with Business Expertise and Reasoning for Automated Requirements Specification and Application Development
摘要
In this study, we present a conversational agent equipped with business expertise and business-oriented reasoning skills. The agent is implemented in a system that is designed to automate the process of developing business applications. The input of the system is a natural language statement expressing a business requirement, and the output is an application that meets end-users’ needs. The system leverages the requirements specification agent presented in this study (a human-machine dialog system equipped with business expertise and reasoning) and a no-code platform called “ONTOMANTICS” that serves as a solution to develop (without code writing) the application, based on specifications returned by the agent. This helps end-users, without development skills, design applications that meet their needs. The agent presented here is responsible for understanding, describing, refining and inferring requirements of end-users. To achieve this, we conduct an experiment to equip the agent with business agility that enables it to adapt to different domains and perform context-specific reasoning.