Combining Convolution and Recursive Neural Networks for Sentiment Analysis

TL;DR

This paper proposes a novel neural network architecture combining Convolution and Recursive Neural Networks for sentence-level sentiment analysis, enhanced by transfer learning, resulting in improved performance over existing models and competitive results on benchmark datasets.

Contribution

It introduces a new combined network architecture and employs transfer learning to enhance sentiment analysis performance, outperforming recent models and matching state-of-the-art results.

Findings

Models outperform recent Convolution and Recursive Neural Networks.

Achieve comparable performance with state-of-the-art systems on Stanford Sentiment Treebank.

Transfer learning improves word embedding quality for sentiment analysis.

Abstract

This paper addresses the problem of sentence-level sentiment analysis. In recent years, Convolution and Recursive Neural Networks have been proven to be effective network architecture for sentence-level sentiment analysis. Nevertheless, each of them has their own potential drawbacks. For alleviating their weaknesses, we combined Convolution and Recursive Neural Networks into a new network architecture. In addition, we employed transfer learning from a large document-level labeled sentiment dataset to improve the word embedding in our models. The resulting models outperform all recent Convolution and Recursive Neural Networks. Beyond that, our models achieve comparable performance with state-of-the-art systems on Stanford Sentiment Treebank.