Evaluating Large Language Models for Document-grounded Response Generation in Information-Seeking Dialogues
This addresses the challenge of evaluating document-grounded dialogue systems for users in social service domains, but it is incremental as it builds on existing datasets and methods.
The paper tackled the problem of using large language models (LLMs) like ChatGPT for generating responses grounded in documents in information-seeking dialogues, finding that ChatGPT variants, despite potential hallucinations, were rated higher by human evaluators than both a shared task winning system and human responses.
In this paper, we investigate the use of large language models (LLMs) like ChatGPT for document-grounded response generation in the context of information-seeking dialogues. For evaluation, we use the MultiDoc2Dial corpus of task-oriented dialogues in four social service domains previously used in the DialDoc 2022 Shared Task. Information-seeking dialogue turns are grounded in multiple documents providing relevant information. We generate dialogue completion responses by prompting a ChatGPT model, using two methods: Chat-Completion and LlamaIndex. ChatCompletion uses knowledge from ChatGPT model pretraining while LlamaIndex also extracts relevant information from documents. Observing that document-grounded response generation via LLMs cannot be adequately assessed by automatic evaluation metrics as they are significantly more verbose, we perform a human evaluation where annotators rate the output of the shared task winning system, the two Chat-GPT variants outputs, and human responses. While both ChatGPT variants are more likely to include information not present in the relevant segments, possibly including a presence of hallucinations, they are rated higher than both the shared task winning system and human responses.