Improving on Q & A Recurrent Neural Networks Using Noun-Tagging

TL;DR

This paper explores enhancing simple QA RNN models by incorporating noun-tagging and hyper-parameter tuning, achieving improved accuracy in entity detection and relation prediction on the SimpleQuestions dataset.

Contribution

The study introduces noun-tagging preprocessing and extensive hyper-parameter tuning to improve QA RNN performance on first-order questions.

Findings

Achieved 0.984 accuracy in entity detection with noun-tagging.

Improved relation prediction accuracy to 0.80 after tuning.

Demonstrated the impact of dataset size on model performance.

Abstract

Often, more time is spent on finding a model that works well, rather than tuning the model and working directly with the dataset. Our research began as an attempt to improve upon a simple Recurrent Neural Network for answering "simple" first-order questions (QA-RNN), developed by Ferhan Ture and Oliver Jojic, from Comcast Labs, using the SimpleQuestions dataset. Their baseline model, a bidirectional, 2-layer LSTM RNN and a GRU RNN, have accuracies of 0.94 and 0.90, for entity detection and relation prediction, respectively. We fine tuned these models by doing substantial hyper-parameter tuning, getting resulting accuracies of 0.70 and 0.80, for entity detection and relation prediction, respectively. An accuracy of 0.984 was obtained on entity detection using a 1-layer LSTM, where preprocessing was done by removing all words not part of a noun chunk from the question. 100% of the dataset was available for relation prediction, but only 20% of the dataset, was available for entity detection, which we believe to be much of the reason for our initial difficulties in replicating their result, despite the fact we were able to improve on their entity detection results.