Complex Factoid Question Answering with a Free-Text Knowledge Graph

We introduce DELFT, a factoid question answering system which combines the nuance and depth of knowledge graph question answering approaches with the broader coverage of free-text. DELFT builds a free-text knowledge graph from Wikipedia, with entities as nodes, and sentences in which entities co-occur as edges. For each question, DELFT finds the subgraph linking question entity nodes to candidate using text sentences as edges, yielding a dense and high coverage semantic graph. A novel graph neural network reasons over the free-text graph—combining evidence on the nodes via information along edge sentences—to select a final answer. Experiments on three question answering datasets show DELFT can answer entity-rich questions better than machine reading based models, BERT-based answer ranking and memory networks with big margins. DELFT’s strong advantage comes from both the high coverage of its free-text knowledge graph—more than doubled that of DBpedia relations—and the novel graph neural network model which conducts accurate structural reasoning on the rich but also noisy free-text evidence.