QA Is the New KR: Question-Answer Pairs as Knowledge Bases

TL;DR

This paper proposes a novel approach to building knowledge bases from text using question-answer pairs, which are modular, user-aligned, and capable of complex reasoning, offering advantages over traditional symbolic KBs.

Contribution

It introduces a new method for creating knowledge bases from text through question generation and entity linking, emphasizing modularity and user relevance.

Findings

Question-answer KBs can answer complex, multi-hop queries.

The approach aligns well with user information needs.

Modular question-answer components facilitate compositional reasoning.

Abstract

In this position paper, we propose a new approach to generating a type of knowledge base (KB) from text, based on question generation and entity linking. We argue that the proposed type of KB has many of the key advantages of a traditional symbolic KB: in particular, it consists of small modular components, which can be combined compositionally to answer complex queries, including relational queries and queries involving "multi-hop" inferences. However, unlike a traditional KB, this information store is well-aligned with common user information needs.