A Discrete Hard EM Approach for Weakly Supervised Question Answering
This addresses the problem of limited supervision in QA for researchers and practitioners, offering a simple yet effective method that is incremental over existing approaches.
The paper tackles weakly supervised question answering by converting tasks into discrete latent variable problems with precomputed solution sets and using a hard EM learning scheme. It shows significant performance improvements, with absolute gains of 2-10% on six QA tasks and state-of-the-art results on five of them.
Many question answering (QA) tasks only provide weak supervision for how the answer should be computed. For example, TriviaQA answers are entities that can be mentioned multiple times in supporting documents, while DROP answers can be computed by deriving many different equations from numbers in the reference text. In this paper, we show it is possible to convert such tasks into discrete latent variable learning problems with a precomputed, task-specific set of possible "solutions" (e.g. different mentions or equations) that contains one correct option. We then develop a hard EM learning scheme that computes gradients relative to the most likely solution at each update. Despite its simplicity, we show that this approach significantly outperforms previous methods on six QA tasks, including absolute gains of 2--10%, and achieves the state-of-the-art on five of them. Using hard updates instead of maximizing marginal likelihood is key to these results as it encourages the model to find the one correct answer, which we show through detailed qualitative analysis.