Pneg: Prompt-based Negative Response Generation for Dialogue Response Selection Task
This work addresses the challenge of improving response selection models in dialogue systems by providing a scalable solution for generating adversarial training data, though it is incremental as it builds on existing methods for negative sample synthesis.
The paper tackles the problem of adversarial responses in retrieval-based dialogue systems by proposing a prompt-based method using a large-scale language model to generate adversarial negative responses, which outperforms existing synthesizing methods and serves as an effective alternative to human annotation.
In retrieval-based dialogue systems, a response selection model acts as a ranker to select the most appropriate response among several candidates. However, such selection models tend to rely on context-response content similarity, which makes models vulnerable to adversarial responses that are semantically similar but not relevant to the dialogue context. Recent studies have shown that leveraging these adversarial responses as negative training samples is useful for improving the discriminating power of the selection model. Nevertheless, collecting human-written adversarial responses is expensive, and existing synthesizing methods often have limited scalability. To overcome these limitations, this paper proposes a simple but efficient method for generating adversarial negative responses leveraging a large-scale language model. Experimental results on dialogue selection tasks show that our method outperforms other methods of synthesizing adversarial negative responses. These results suggest that our method can be an effective alternative to human annotators in generating adversarial responses. Our dataset and generation code is available at https://github.com/leenw23/generating-negatives-by-gpt3.