Compositional Sentence Representation from Character within Large Context Text

TL;DR

This paper introduces a Hierarchical Composition Recurrent Network (HCRN) that constructs sentence representations from characters to words to sentences, effectively capturing semantics and inter-sentence dependencies, leading to state-of-the-art dialogue act classification performance.

Contribution

The paper presents a novel hierarchical model that integrates character-level composition and inter-sentence dependencies, addressing data sparsity and improving semantic understanding.

Findings

Achieved 22.7% error rate on SWBD-DAMSL dataset

Effectively captures implicit and explicit sentence semantics

Outperforms previous models in dialogue act classification

Abstract

This paper describes a Hierarchical Composition Recurrent Network (HCRN) consisting of a 3-level hierarchy of compositional models: character, word and sentence. This model is designed to overcome two problems of representing a sentence on the basis of a constituent word sequence. The first is a data-sparsity problem in word embedding, and the other is a no usage of inter-sentence dependency. In the HCRN, word representations are built from characters, thus resolving the data-sparsity problem, and inter-sentence dependency is embedded into sentence representation at the level of sentence composition. We adopt a hierarchy-wise learning scheme in order to alleviate the optimization difficulties of learning deep hierarchical recurrent network in end-to-end fashion. The HCRN was quantitatively and qualitatively evaluated on a dialogue act classification task. Especially, sentence representations with an inter-sentence dependency are able to capture both implicit and explicit semantics of sentence, significantly improving performance. In the end, the HCRN achieved state-of-the-art performance with a test error rate of 22.7% for dialogue act classification on the SWBD-DAMSL database.