QAMPARI: An Open-domain Question Answering Benchmark for Questions with Many Answers from Multiple Paragraphs
This addresses the need for benchmarks that handle multi-answer questions in open-domain QA, though it is incremental as it extends existing single-answer benchmarks.
The authors tackled the problem of open-domain question answering for questions with multiple answers spread across many paragraphs by introducing QAMPARI, a new benchmark, and found that existing models achieve only a 32.8 F1 score, highlighting its difficulty.
Existing benchmarks for open-domain question answering (ODQA) typically focus on questions whose answers can be extracted from a single paragraph. By contrast, many natural questions, such as "What players were drafted by the Brooklyn Nets?" have a list of answers. Answering such questions requires retrieving and reading from many passages, in a large corpus. We introduce QAMPARI, an ODQA benchmark, where question answers are lists of entities, spread across many paragraphs. We created QAMPARI by (a) generating questions with multiple answers from Wikipedia's knowledge graph and tables, (b) automatically pairing answers with supporting evidence in Wikipedia paragraphs, and (c) manually paraphrasing questions and validating each answer. We train ODQA models from the retrieve-and-read family and find that QAMPARI is challenging in terms of both passage retrieval and answer generation, reaching an F1 score of 32.8 at best. Our results highlight the need for developing ODQA models that handle a broad range of question types, including single and multi-answer questions.