Beyond I.I.D.: Three Levels of Generalization for Question Answering on Knowledge Bases

Abstract

Existing studies on question answering on knowledge bases (KBQA) mainly operate with the standard i.i.d assumption, i.e., training distribution over questions is the same as the test distribution. However, i.i.d may be neither reasonably achievable nor desirable on large-scale KBs because 1) true user distribution is hard to capture and 2) randomly sample training examples from the enormous space would be highly data-inefficient. Instead, we suggest that KBQA models should have three levels of built-in generalization: i.i.d, compositional, and zero-shot. To facilitate the development of KBQA models with stronger generalization, we construct and release a new large-scale, high-quality dataset with 64,331 questions, GrailQA, and provide evaluation settings for all three levels of generalization. In addition, we propose a novel BERT-based KBQA model. The combination of our dataset and model enables us to thoroughly examine and demonstrate, for the first time, the key role of pre-trained contextual embeddings like BERT in the generalization of KBQA.

arXiv commentAccepted to TheWebConf 2021 (previously WWW)

Cited by in corpus (7)

• QA Dataset Explosion: A Taxonomy of NLP Resources for Question Answering and Reading Comprehension
• LEGO-GraphRAG: Modularizing Graph-based Retrieval-Augmented Generation for Design Space Exploration
• Would You Ask it that Way? Measuring and Improving Question Naturalness for Knowledge Graph Question Answering
• In-Context Learning for Knowledge Base Question Answering for Unmanned Systems based on Large Language Models
• Lexicalization Is All You Need: Examining the Impact of Lexical Knowledge in a Compositional QALD System
• The benefits of query-based KGQA systems for complex and temporal questions in LLM era
• BioGraphletQA: Knowledge-Anchored Generation of Complex QA Datasets