用合成数据训练对话问答系统，让模型理解上下文并引用文档。
Building Open-Retrieval Conversational Question Answering Systems by Generating Synthetic Data and Decontextualizing User Questions
- 利用组织文本自动生成真实对话数据，包含问答对和文档依据。
- 合成数据训练出能去上下文化问题的模型，适配旧检索器。
- 适合构建企业级对话问答系统，无需人工标注数据。
我们研究开放域对话问答（OR-CONVQA），要求系统响应既要理解对话历史，又要基于检索到的文档内容。领域专用的OR-CONVQA训练数据对实际应用至关重要，但难以获取。本文提出一种流水线方法，利用组织中丰富的纯文本文档（如产品文档）自动生成具有标注的逼真对话。与真实的人工标注数据类似，生成的对话包含对话内问答对、去上下文化的用户问题版本（无指代词），以及系统响应所依据的主张（即文档中的关键句子）。我们展示了如何用这些合成对话训练高效的问句重写模型，将用户问题去上下文化，从而兼容不支持对话的旧式检索器。随后，将去上下文化的问题与检索信息输入大模型，生成最终回答。
原文摘要 · Abstract (English)
We consider open-retrieval conversational question answering (OR-CONVQA), an extension of question answering where system responses need to be (i) aware of dialog history and (ii) grounded in documents (or document fragments) retrieved per question. Domain-specific OR-CONVQA training datasets are crucial for real-world applications, but hard to obtain. We propose a pipeline that capitalizes on the abundance of plain text documents in organizations (e.g., product documentation) to automatically produce realistic OR-CONVQA dialogs with annotations. Similarly to real-world humanannotated OR-CONVQA datasets, we generate in-dialog question-answer pairs, self-contained (decontextualized, e.g., no referring expressions) versions of user questions, and propositions (sentences expressing prominent information from the documents) the system responses are grounded in. We show how the synthetic dialogs can be used to train efficient question rewriters that decontextualize user questions, allowing existing dialog-unaware retrievers to be utilized. The retrieved information and the decontextualized question are then passed on to an LLM that generates the system's response.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。