Faithful Embeddings for Knowledge Base Queries
This addresses the issue of unreliable query answers in knowledge base systems for users needing accurate inference, though it is incremental as it builds on existing QE techniques.
The paper tackles the problem that query embedding (QE) systems can produce answers inconsistent with deductive reasoning even when generalization is not needed, by introducing a novel QE method that is more faithful to deductive reasoning. This leads to better performance on complex queries to incomplete knowledge bases and substantial improvements in a neural question-answering system over the state-of-the-art.
The deductive closure of an ideal knowledge base (KB) contains exactly the logical queries that the KB can answer. However, in practice KBs are both incomplete and over-specified, failing to answer some queries that have real-world answers. \emph{Query embedding} (QE) techniques have been recently proposed where KB entities and KB queries are represented jointly in an embedding space, supporting relaxation and generalization in KB inference. However, experiments in this paper show that QE systems may disagree with deductive reasoning on answers that do not require generalization or relaxation. We address this problem with a novel QE method that is more faithful to deductive reasoning, and show that this leads to better performance on complex queries to incomplete KBs. Finally we show that inserting this new QE module into a neural question-answering system leads to substantial improvements over the state-of-the-art.