Question Answering with Subgraph Embeddings
This addresses question answering for users needing automated responses from structured data, but it is incremental as it builds on existing embedding and training methods.
The paper tackles the problem of answering questions across diverse topics from a knowledge base by learning embeddings for words and knowledge base elements to score questions against answers, achieving competitive results on a benchmark.
This paper presents a system which learns to answer questions on a broad range of topics from a knowledge base using few hand-crafted features. Our model learns low-dimensional embeddings of words and knowledge base constituents; these representations are used to score natural language questions against candidate answers. Training our system using pairs of questions and structured representations of their answers, and pairs of question paraphrases, yields competitive results on a competitive benchmark of the literature.