Joint Multitask Learning for Community Question Answering Using Task-Specific Embeddings
This work addresses the problem of improving question and answer retrieval in community forums, which is incremental as it combines existing DNN and CRF methods in a multitask setting.
The paper tackled the problem of jointly performing related question retrieval and answer retrieval in community question answering by using deep neural networks to learn task-specific embeddings and incorporating them into a conditional random field model for multitask learning. The result showed that the CRF model significantly and consistently improved performance over DNNs alone across various evaluation metrics.
We address jointly two important tasks for Question Answering in community forums: given a new question, (i) find related existing questions, and (ii) find relevant answers to this new question. We further use an auxiliary task to complement the previous two, i.e., (iii) find good answers with respect to the thread question in a question-comment thread. We use deep neural networks (DNNs) to learn meaningful task-specific embeddings, which we then incorporate into a conditional random field (CRF) model for the multitask setting, performing joint learning over a complex graph structure. While DNNs alone achieve competitive results when trained to produce the embeddings, the CRF, which makes use of the embeddings and the dependencies between the tasks, improves the results significantly and consistently across a variety of evaluation metrics, thus showing the complementarity of DNNs and structured learning.