基于 RAE+Dropout 相结合的微博情感分析

文本情感分析是近年自然语言处理领域的研究热点之一，其中微博情感分析受到了学术界和企业界的广泛关注。微博情感分析是指对用户针对某一事件发表的言论进行正向、负向和中性情感的判定。本文在分析了标准RAE 模型缺点的基础上，提出了一种基于RAE+Dropout 的联合模型。该模型利用Dropout 技术有效地预防过拟合问题的发生，同时也提高了模型训练速度。RAE+Dropout 模型与RAE+ 词性选择模型、标准RAE 模型以及SVM 模型的对比实验结果表明：RAE+Dropout 模型的准确率和F1 值属于最优，比标准RAE 模型的准确值和F1 值高出0.82% 和0.64%，尤其是在高维词语向量中RAE+Dropout 模型的效果更加明显。

Micro-blog Emotion Analysis Based on RAE + Dropout

Recently, text emotion analysis is one of the most popular research hotpots in the field of natural language processing, Furthermore, micro-blog emotion analysis has received extensive attention in academia and industry. Micro-blog emotion analysis aims to judge the positive, negative and neutral emotion of the speech delivered by the user for an event. In this paper, firstly, the shortcomings of the original model were illustrated according to the standard RAE model, and then an improved model, which is a joint model of RAE + Dropout was put forward. This model aimed to prevent the problem of over-fitting as well as improve the speed of model training by Dropout technology. At last, The RAE + Dropout model was compared with the RAE + POS model, the standard RAE model and the SVM model respectively. The experiment results indicated that: RAE + Dropout model has the best accuracy and F1 value, which are 0.82% and 0.64% up on the results of the standard RAE model, especially the RAE + Dropout model has an obvious advantage in the high-dimensional word vector representation.