Applying Knowledge Management to Support Artificial Intelligence Chatbot Applications

As the diversity and complexity of Artificial Intelligence (AI) systems increase, there is a growing need for advanced knowledge representation methods to enhance decision-making capabilities. Existing research indicates a gap between AI and Knowledge Management (KM), emphasizing the necessity of coordinating learning and knowledge creation processes between humans and machines. Despite the widespread use of generative AI, as seen through the growing popularity of conversational AI tools like chatbots powered by Large Language Models in recent years, the absence of a theoretical framework for effectively managing the knowledge they generate could mean missing out on significant opportunities. This work seeks to bridge this gap between KM and IA through an integrated framework that aims to apply KM to support IA chatbot applications, adapted from the Internet of Everything Integrated Knowledge Management Model (IoE IKM Model). The IoE IKM Model’s original goal is to support knowledge creation in IoE applications, but here, we show how it can be adapted to bring KM to the context of AI. We accomplish this by explaining the development process of the IoE IKM Model, identifying shared aspects between IoE and AI general applications, and adapting necessary elements to establish our integrated KM framework tailored for supporting AI chatbot applications. The resulting framework is then discussed, and examples of how it can be applied to enhance human interaction with a chatbot, namely Open AI's ChatGPT. Research has been conducted to demonstrate the advantages of applying AI in KM. However, we aim to take a different approach by showing how KM can contribute to AI applications. We expect this work to be helpful for those whose professional activities may involve the usage of AI systems by providing them with the necessary tools to manage the knowledge generated by these same AI systems and by offering a Knowledge Manager’s perspective on how to boost human-machine interaction.