Bag-of-Vector Embeddings of Dependency Graphs for Semantic Induction
This addresses a bottleneck in NLP for representing complex linguistic structures, though it is incremental as it builds on existing vector-space models.
The paper tackles the problem of generalizing vector-space models from fixed-length word vectors to arbitrary linguistic structures by proposing bag-of-vector embeddings for dependency graphs, achieving competitive results on Semantic Textual Similarity and Natural Language Inference tasks.
Vector-space models, from word embeddings to neural network parsers, have many advantages for NLP. But how to generalise from fixed-length word vectors to a vector space for arbitrary linguistic structures is still unclear. In this paper we propose bag-of-vector embeddings of arbitrary linguistic graphs. A bag-of-vector space is the minimal nonparametric extension of a vector space, allowing the representation to grow with the size of the graph, but not tying the representation to any specific tree or graph structure. We propose efficient training and inference algorithms based on tensor factorisation for embedding arbitrary graphs in a bag-of-vector space. We demonstrate the usefulness of this representation by training bag-of-vector embeddings of dependency graphs and evaluating them on unsupervised semantic induction for the Semantic Textual Similarity and Natural Language Inference tasks.