Complex Factoid Question Answering with a Free-Text Knowledge Graph
This addresses the problem of improving accuracy and coverage in factoid question answering for applications like search engines or AI assistants, though it is an incremental advance over prior knowledge graph and text-based methods.
The paper tackles factoid question answering by introducing DELFT, a system that constructs a free-text knowledge graph from Wikipedia and uses a novel graph neural network to reason over it, achieving better performance on entity-rich questions than existing models like BERT-based ranking and memory networks, with more than double the coverage of DBpedia relations.
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 candidates using text sentences as edges, creating 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. DELFT's advantage comes from both the high coverage of its free-text knowledge graph-more than double that of dbpedia relations-and the novel graph neural network which reasons on the rich but noisy free-text evidence.