Semantic Sentence Matching with Densely-connected Recurrent and Co-attentive Information

29 May 2018  ·  Seonhoon Kim, Inho Kang, Nojun Kwak ·

Sentence matching is widely used in various natural language tasks such as natural language inference, paraphrase identification, and question answering. For these tasks, understanding logical and semantic relationship between two sentences is required but it is yet challenging. Although attention mechanism is useful to capture the semantic relationship and to properly align the elements of two sentences, previous methods of attention mechanism simply use a summation operation which does not retain original features enough. Inspired by DenseNet, a densely connected convolutional network, we propose a densely-connected co-attentive recurrent neural network, each layer of which uses concatenated information of attentive features as well as hidden features of all the preceding recurrent layers. It enables preserving the original and the co-attentive feature information from the bottommost word embedding layer to the uppermost recurrent layer. To alleviate the problem of an ever-increasing size of feature vectors due to dense concatenation operations, we also propose to use an autoencoder after dense concatenation. We evaluate our proposed architecture on highly competitive benchmark datasets related to sentence matching. Experimental results show that our architecture, which retains recurrent and attentive features, achieves state-of-the-art performances for most of the tasks.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Natural Language Inference SNLI Densely-Connected Recurrent and Co-Attentive Network % Test Accuracy 88.9 # 22
% Train Accuracy 93.1 # 25
Parameters 6.7m # 4
Natural Language Inference SNLI Densely-Connected Recurrent and Co-Attentive Network Ensemble % Test Accuracy 90.1 # 11
% Train Accuracy 95.0 # 12
Parameters 53.3m # 4
Natural Language Inference SNLI Densely-Connected Recurrent and Co-Attentive Network (encoder) % Test Accuracy 86.5 # 55
% Train Accuracy 91.4 # 36
Parameters 5.6m # 4

Methods