Adapted GooLeNet for Answer Selection
Huang J. Jie, Qi Zhang, Long Qin · 2018
Semantic matching is of central significance to the answer selection task which aims to select correct answers for a given question from a candidate answer pool. A useful method is to employ neural networks to model sentence by generating sentences representations and then measuring their distance. In this work, we introduce an effective architecture --Adapted GooLeNet (AG)-- into the answer selection task for sentence modeling. This architecture can capture more levels of language granularities in parallel, because of the various sizes of filters comparing with single-layer CNN and multi-layer CNN. The empirical study on three various benchmark tasks of answer selection demonstrates that capturing sentence features on different levels of granularities benefit sentence modeling by utilizing AG, comparing with single-layer CNNs, multi-layer CNNs and biLSTM.