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

We propose a new knowledge representation (KR) based on bases (KBs) derived from text, question generation and entity linking. argue that the proposed type of KB has many key advantages traditional symbolic KB: in particular, it consists small modular components, which can be combined compositionally to answer complex queries, including relational queries involving ``multi-hop'' inferences. However, unlike KB, this information store is well-aligned with common user needs. present one such called QEDB, give qualitative evidence atomic components are high-quality meaningful, ways similar triples KB. also show experimentally questions reflective typical more easily answered QEDB than