Abstract
Multi-hop knowledge graph (KG) reasoning has been widely studied in recent years to provide interpretable predictions on missing links. Given a triple query, multi-hop reasoning task aims to give an evidential path that indicates the inference process. Most previous works use reinforcement learning (RL) based method that learns to navigate the path towards the target entity. However, these methods suffer from slow and poor convergence, and they may fail to infer a certain path when there is a missing edge along the path. Here we present SQUIRE, the first Sequence-to-sequence based multi-hop reasoning framework, which utilizes an encoder-decoder structure to translate the query to a path. Our model design brings about two benefits: (1) It can learn and predict in an end-to-end fashion, which gives better and faster convergence; (2) Our model does not rely on existing edges to generate the path, and has the flexibility to complete missing edges along the path, especially in sparse KGs. Experiments on standard and sparse KGs show that our approach yields significant improvement over prior methods, while converging 4x-7x faster.
Abstract (translated)
URL
https://arxiv.org/abs/2201.06206
