Effective shared representations with Multitask Learning for Community Question Answering

An important asset of using Deep Neu- ral Networks (DNNs) for text applica- tions is their ability to automatically engi- neer features. Unfortunately, DNNs usu- ally require a lot of training data, espe- cially for high-level semantic tasks such as community Question Answering (cQA). In this paper, we tackle the problem of data scarcity by learning the target DNN together with two auxiliary tasks in a mul- titask learning setting. We exploit the strong semantic connection between se- lection of comments relevant to (i) new questions and (ii) forum questions. This enables a global representation for com- ments, new and previous questions. The experiments of our model on a SemEval challenge dataset for cQA show a 20% rel- ative improvement over standard DNNs.