基于Word2Vec文本情感研究与分析.pdfVIP

Computer Science and Application 计算机科学与应用, 2017, 7(10), 944-950 Published Online October 2017 in Hans. /journal/csa /10.12677/csa.2017.710107 Research and Analysis of Textual Emotion Based on Word2Vec Guanghua Yin School of Computer Science, Zhongyuan University of Technology, Zhengzhou Henan th th th Received: Sep. 30 , 2017; accepted: Oct. 13 , 2017; published: Oct. 19 , 2017 Abstract With the development of artificial intelligence, machine learning development, compared with traditional neural network and other non-linear decision and modeling theory, support vector is used in the field of text information processing to solve the classification problem, because of the simple structure and complete theories. This paper puts forward the research on the emotion classification of the depth learning text in view of the intricacies of the massive text emotion data, the inability to grasp the positive and negative information of the text accurately and accurately. Firstly, we introduce the idea of text emotion classification. Then we introduce the text emotion and TF-IDF weight calculation, and improve the TF-IDF weight and mediate the depth of learning Word2Ve word vector + LIBSVM model to train the Internet text data. Finally, the accuracy of the classification reached 92.28%. Keywords Machine Learning, LIBSVM, Emotional Characteristics, Word2Vec, Emotional Classification 基于Word2Vec文本情感研究与分析 尹光花 中原工学院，计算机学院，河南 郑州 收稿日期：2017年9月30 日；录用日期：2017年10月13 日；发布日期：2017年10月19 日 摘 要 随着人工智能、机器学习发展，支持向量与传统的神经网络等非线性判定和建模理论相比，因结构简单， 文章引用: 尹光花. 基于Word2Vec 文本情感研究与分析[J]. 计算机科学与应用, 2017, 7(10): 944-950. DOI: 10.12677/csa.2017.710107