QUADRo: Dataset and Models for QUestion-Answer Database Retrieval
This work addresses the challenge of efficient and accurate open-domain question answering for applications like FAQs or forums, though it is incremental as it builds on existing retrieval paradigms.
The paper tackles the problem of scaling question-answering systems by reusing previously answered questions from a large database, achieving competitive performance with standard methods like unstructured document retrieval. They built a dataset of 6.3 million question/answer pairs and showed that Transformer-based models using both questions and answers outperform question-only models in neural search and reranking.
An effective paradigm for building Automated Question Answering systems is the re-use of previously answered questions, e.g., for FAQs or forum applications. Given a database (DB) of question/answer (q/a) pairs, it is possible to answer a target question by scanning the DB for similar questions. In this paper, we scale this approach to open domain, making it competitive with other standard methods, e.g., unstructured document or graph based. For this purpose, we (i) build a large scale DB of 6.3M q/a pairs, using public questions, (ii) design a new system based on neural IR and a q/a pair reranker, and (iii) construct training and test data to perform comparative experiments with our models. We demonstrate that Transformer-based models using (q,a) pairs outperform models only based on question representation, for both neural search and reranking. Additionally, we show that our DB-based approach is competitive with Web-based methods, i.e., a QA system built on top the BING search engine, demonstrating the challenge of finding relevant information. Finally, we make our data and models available for future research.