Towards Reading Comprehension for Long Documents

Authors: Yuanxing Zhang, Yangbin Zhang, Kaigui Bian, Xiaoming Li

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on the modified SQu AD dataset show that our proposed model outperforms existing reading comprehension models by at least 20% regarding exact match (EM), F1 and the proportion of identified paragraphs which are exactly the short paragraphs where the original answers locate.
Researcher Affiliation Academia Yuanxing Zhang, Yangbin Zhang, Kaigui Bian and Xiaoming Li Peking University, China {longo,yangbin zhang,bkg,lxm}@pku.edu.cn
Pseudocode No The paper describes the model architecture and components but does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described, nor does it explicitly state that code is released or available.
Open Datasets Yes We use the SQuAD v1.1 dataset to conduct the experiments.
Dataset Splits Yes We split the training set into a training set (81,398 tuples) and a validation set (4,285 tuples).
Hardware Specification Yes The experiments run on two GTX1080Ti
Software Dependencies No The paper mentions GloVe and Adam optimization but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes We use the pre-trained GloVe [Pennington et al., 2014] parameters with 300-dimension embeddings to initialize the model. ... the hidden size of every recurrent network involved in our model is fixed as 150 in single direction. ... The learning rate is dynamically updated according to Adam optimization [Kingma and Ba, 2014]. Besides, in order to avoid overfitting or out of memory, the gradient descent is updated by minibatch of 32 instances.