Conversation Intention Perception Based on Knowledge Base

Web Intelligence is gaining its growth in a rapid speed. The notion of wisdom, which is considered as the next paradigm shift of WI, has become a hot research topic in recent years. The basic application of wisdom is making a short conversation in an interactive and understandable way based on the huge web resources. However, current conversation system normally applies the recognition of semantic similarities in the prepared database, neglecting the true intention hiding in the expression. In this paper, we present a model based on the medical Q&A knowledge base to overcome this challenge. The knowledge base includes three parts: disease entity, medicine, properties. A simple graph path algorithm based on words direction and relation weight adjustment is used to realize conversation intention perception. The experimental results show that this method can effectively perceive types of intention. This method can also be applied in deep understanding of other intelligent systems such as classifications and text mining.