Comparing Attention-based Convolutional and Recurrent Neural Networks: Success and Limitations in Machine Reading Comprehension
arXiv:1808.08744
Abstract
We propose a machine reading comprehension model based on the compare-aggregate framework with two-staged attention that achieves state-of-the-art results on the MovieQA question answering dataset. To investigate the limitations of our model as well as the behavioral difference between convolutional and recurrent neural networks, we generate adversarial examples to confuse the model and compare to human performance. Furthermore, we assess the generalizability of our model by analyzing its differences to human inference,
arXiv commentCoNLL 2018
References in corpus (7)
- Efficient Estimation of Word Representations in Vector Space
- Natural Language Processing (almost) from Scratch
- Convolutional Neural Network Architectures for Matching Natural Language Sentences
- Comparative Study of CNN and RNN for Natural Language Processing
- A C-LSTM Neural Network for Text Classification
- Adversarial Examples: Attacks and Defenses for Deep Learning
- A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task