Machine Comprehension using SQuAD and Deep Learning
نویسندگان
چکیده
There are many state-of-the-art deep learning architectures that have been used to develop reading comprehension systems using the Stanford Question Answering Dataset (SQuAD). In this work, we provide an overview of some of the state-ofthe-art models and compare their performance across several variations of encoder and decoder architectures and different attention mechanisms. We take inspiration from these state-of-the-art models and develop our own hybrid architectures, which have performed very similarly to previous models in the literature such as Match-LSTM and Dynamic Coattention when excluding extensive hyperparameter searches. Across all the architectures experimented with, our custom Stackson-Stacks (SoS) model yielded the best results of 0.537 F1 and 0.407 EM on the SQuAD test set.
منابع مشابه
SQuAD Reading Comprehension with Coattention
Reading comprehension is an important task in NLP, which involves teaching a machine to understand text enough to answer questions. The Stanford Question Answering Dataset (SQuAD) is a dataset consisting of 100,000 question-context-answer datapoints. Here, deep learning methods are used to answer questions based on context data. A model based on the Attentive Reader [1,2] model is used as a bas...
متن کاملSQuAD reading comprehension deep learning model
We introduce a neural network model for reading comprehension using the SQuAD dataset. Our model is composed of a Dynamic Coattention Network encoder (Xiong et al. [2016]) and a novel decoder designed for runtime minimization. Our model obtained an F1 score of 52.283 when tested on the SQuAD dev set, and an exact-match score of 38.723.
متن کاملA new model for Machine Comprehension via multi-perspective context matching and bidrectional attention flow
To answer a question about a context paragraph, there needs to be a complex model for interactions between these two. Previous Machine Comprehension (MC) where either not large enough to train end-to-end deep neural networks, or not hard to learn. Recently, after the release of SQuAD dataset dataset, several adept models have been proposed for the task of MC. In this work we try to combine the ...
متن کاملMulti-Perspective Context Matching for Machine Comprehension
Previous machine comprehension (MC) datasets are either too small to train endto-end deep learning models, or not difficult enough to evaluate the ability of current MC techniques. The newly released SQuAD dataset alleviates these limitations, and gives us a chance to develop more realistic MC models. Based on this dataset, we propose a Multi-Perspective Context Matching (MPCM) model, which is ...
متن کاملAssignment 4: Question Answering on the SQuAD Dataset with Part-of-speech Tagging
This research applies deep learning with bi-LSTMs to train a model that responds to queries on the Stanford Question Answering Dataset (SQuAD). The model design was motivated by Wang et. al.’s December 2016 IBM Research paper on multi-perspective context matching for machine comprehension. Using TensorFlow, we implemented a multi-layer neural network architecture utilizing bi-directional LSTMs ...
متن کامل