Sanity Check: A Strong Alignment and Information Retrieval Baseline for Question Answering

TL;DR

This paper introduces a simple, unsupervised baseline for question answering that uses novel alignment techniques, outperforming many existing methods and approaching state-of-the-art results across multiple datasets.

Contribution

The paper presents a novel unsupervised alignment and information retrieval method with one-to-many and negative alignment, providing a strong baseline for QA tasks.

Findings

Outperforms conventional baselines and many supervised RNNs.

Achieves high P@1 and MAP scores on multiple QA datasets.

Approaches state-of-the-art performance with minimal hyperparameters.

Abstract

While increasingly complex approaches to question answering (QA) have been proposed, the true gain of these systems, particularly with respect to their expensive training requirements, can be inflated when they are not compared to adequate baselines. Here we propose an unsupervised, simple, and fast alignment and information retrieval baseline that incorporates two novel contributions: a \textit{one-to-many alignment} between query and document terms and \textit{negative alignment} as a proxy for discriminative information. Our approach not only outperforms all conventional baselines as well as many supervised recurrent neural networks, but also approaches the state of the art for supervised systems on three QA datasets. With only three hyperparameters, we achieve 47\% P@1 on an 8th grade Science QA dataset, 32.9\% P@1 on a Yahoo! answers QA dataset and 64\% MAP on WikiQA. We also achieve 26.56\% and 58.36\% on ARC challenge and easy dataset respectively. In addition to including the additional ARC results in this version of the paper, for the ARC easy set only we also experimented with one additional parameter -- number of justifications retrieved.