Chinese Microblog Sentiment Classification Based on Deep Belief Nets with Extended Multi-Modality Features

Abstract

This paper presents a DBN (deep belief nets) model and a multi-modality feature extraction method to extend features' dimensionalities of short text for Chinese micro blogging sentiment classification. Besides traditional features sets for document classification, comments for certain posts are also extracted as part of the micro blogging features according to the relationship between commenters and posters though constructing micro blogging social network as input information. Then, the integration of the above modality features is combined and represented as input vector for DBN. In this paper, a DBN model, which is stacked with several layers of RBM (Restricted Boltzmann Machine), is implemented to initialize the structure of neural network. The RBM layers can take probability distribution samples of original data to learn hidden structures for better feature representation. A Class RBM (Classification RBM) layer, which is stacked on top of several RBM layers, is adapted to achieve the final sentiment classification. The results demonstrate that, with proper structure and parameter, the performance of the proposed deep learning method on sentiment classification is better than state of the art surface learning models such as SVM or NB, which proves that DBN is suitable for short-length document classification with the proposed feature dimensionality extension method.