Question Answering with Subgraph Embeddings

TL;DR

This paper introduces a system that uses subgraph embeddings to answer questions across various topics by learning low-dimensional representations, achieving competitive results with minimal hand-crafted features.

Contribution

The paper proposes a novel approach that learns subgraph embeddings for question answering, reducing reliance on manual feature engineering and improving performance.

Findings

Achieved competitive benchmark results

Learned effective low-dimensional embeddings

Reduced need for hand-crafted features

Abstract

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.