Shallow Discourse Parsing Using Distributed Argument Representations and Bayesian Optimization

TL;DR

This paper presents a neural network approach for discourse relation sense classification using LSTM-derived argument representations and Bayesian optimization for hyperparameter tuning.

Contribution

It introduces a method combining LSTM-based argument representations with surface features and employs Bayesian optimization to determine neural network architecture.

Findings

Effective discourse relation classification achieved

Bayesian hyperparameter search improves model performance

LSTM representations enhance argument understanding

Abstract

This paper describes the Georgia Tech team's approach to the CoNLL-2016 supplementary evaluation on discourse relation sense classification. We use long short-term memories (LSTM) to induce distributed representations of each argument, and then combine these representations with surface features in a neural network. The architecture of the neural network is determined by Bayesian hyperparameter search.