Compositional Vector Space Models for Knowledge Base Inference.
Arvind Neelakantan, Benjamin Roth, Andrew McCallum · 2015
Knowledge base (KB) completion adds new facts to a KB by making inferences from existing facts, for example by infer-ring with high likelihood nationality(X,Y) from bornIn(X,Y). Most previous methods infer simple one-hop relational synonyms like this, or use as evidence a multi-hop re-lational path treated as an atomic feature, like bornIn(X,Z) → containedIn(Z,Y). This paper presents an approach that reasons about conjunctions of multi-hop relations non-atomically, composing the implica-tions of a path using a recurrent neural network (RNN) that takes as inputs vec-tor embeddings of the binary relation in the path. Not only does this allow us to generalize to paths unseen at training time, but also, with a single high-capacity RNN, to predict new relation types not seen when the compositional model was trained (zero-shot learning). We assem-ble a new dataset of over 52M relational triples, and show that our method im-proves over a traditional classifier by 11%, and a method leveraging pre-trained em-beddings by 7%. 1