Full-Time Supervision based Bidirectional RNN for Factoid Question Answering
This work addresses a specific bottleneck in QA models for researchers, though it appears incremental as it builds on existing BRNN methods.
The paper tackled the problem of information loss in bidirectional RNNs for factoid question answering by proposing a full-time supervision model that applies supervision at every time step, achieving state-of-the-art accuracy in experiments.
Recently, bidirectional recurrent neural network (BRNN) has been widely used for question answering (QA) tasks with promising performance. However, most existing BRNN models extract the information of questions and answers by directly using a pooling operation to generate the representation for loss or similarity calculation. Hence, these existing models don't put supervision (loss or similarity calculation) at every time step, which will lose some useful information. In this paper, we propose a novel BRNN model called full-time supervision based BRNN (FTS-BRNN), which can put supervision at every time step. Experiments on the factoid QA task show that our FTS-BRNN can outperform other baselines to achieve the state-of-the-art accuracy.