Learning to Acquire Knowledge from a Search Engine for Dialogue Response Generation

Anonymous

Abstract: Knowledge-aided dialogue response generation aims at augmenting chatbots with relevant external knowledge in the hope of generating more informative responses.The majority of previous work assumes that the relevant knowledge is given as input or retrieved from a static pool of knowledge. However, this assumption violates the real-world situation, where knowledge is continually updated and a chatbot has to \emph{dynamically} retrieve useful knowledge.In this paper, we propose a dialogue model that can access the vast and dynamic information from any search engine for response generation. To this end, we design a query producer that generates queries from a dialogue context to interact with a search engine. The query producer is trained without any human annotation of gold queries, making it easily transferable to other domains and search engines. More specifically, we design a reinforcement learning algorithm to train the query producer, where rewards are obtained by comparing retrieved articles and gold responses. Experiments show that our query producer can achieve R@$1$ and R@$5$ rates of 62.4\% and 74.8\% for retrieving gold knowledge, and the overall model generates better responses over a strong BART (Lewis et al., 2020) model and other typical baselines.