Question Answering Retrieval Method for Knowledge-Based with Dynamic Programming Algorithm

This paper provides a Chinese question and answer retrieval method based on dynamic programming algorithm. This paper aims to unify the question and answer process, which was originally divided into three sub steps, into the same process, and integrate the idea of state machine in formal language to model the entity recognition process and the question and answer retrieval process with the same model. Entity recognition is modeled as a transfer process from the initial state 0 to state 1 of the system, The question and answer retrieval process is modeled as a transition from state 1 to state 2. To ensure the accuracy and breadth of answers, and to unify the whole question and answer process from a macro perspective. In addition, this paper models the transfer process of the target entity in the question answering process as the state transfer process in the state machine, models the transit entity as the intermediate state, and models the final answer as the termination state. On this basis, the dynamic programming algorithm is used to support the overall decision making of the transition process between all states, while ensuring the operating efficiency of the system.