[目的/意义]通过基于Word2vec的文本情感分析技术对某显示器品牌的产品与服务的在线评论进行分析，研究消费者的品牌认知和品牌口碑，为管理者建立更科学的品牌管理体系提供有针对性的建议。[方法/过程]首先利用自然语言处理技术，对评论语料库进行预处理，结合深度学习的Word2vec词向量技术构建产品特征词和情感词词库，进一步构造情感概念对进行情感评分，并将其用于分析品牌产品特定特征的用户情感。[结果/结论]通过Word2vec构建的情感词典相较于传统方法（例如一般的情绪词表）进行情感分析，在情感分析的准确率上有所提高，再结合有效的情感概念对构造与情感评分，可以有效地理解用户的品牌认知。
[Purpose/significance] Through the text sentiment analysis technology of the display products' online reviews based on Word2vec, this paper studies the brand awareness and brand reputation of consumers, and provides the feasible suggestions for managers to establish a more scientific brand management system.[Method/process] This paper used the natural language processing technology to preprocess the comments corpus. It combined the deep learning Word2vec technology to build the product key and emotional word dictionary to analyze the characteristic of the brand specific emotions.[Result/conclusion] Compared with the conventional methods (e.g. general emotional vocabulary), the sentiment dictionary constructed by Word2vec improves the accuracy of sentiment analysis, and effectively helps us understand the user's brand awareness with the effective construction of the sentiment concept pair and the sentiment score.
[1] MUNIZ A M, O'GUINN T C. Brand community[J]. Journal of consumer research, 1998, 27(4):412-432.
[2] COLLOBERT R, WESTON J, KARLEN M. Natural language processing (almost) from scratch[J]. Journal of machine learning research, 2011, 12(1):2493-2537.
[3] 蓝天广. 电子商务产品在线评论的细粒度情感强度分析[D]. 北京:北京邮电大学, 2015.
[4] 杨桂芳. 网络评论挖掘关键技术与方法研究概述[J]. 计算机光盘软件与应用,2014(22):122-123.
[5] HU M, LIU B. Mining and summarizing customer reviews[C]//Tenth ACM SIGKDD international conference on knowledge discovery and data mining, Seattle, Washington, Usa, August. New York:ACM, 2004:168-177.
[6] WANG B, WANG H. Bootstrapping both product properties and opinion words from chinese reviews with cross-training[C]//Web intelligence, IEEE/WIC/ACM international conference on. Los Alamitos:IEEE Computer Society, 2007:259-262.
[7] 徐琳宏, 林鸿飞, 潘宇,等. 情感词汇本体的构造[J]. 情报学报, 2008, 27(2):180-185.
[8] 谢丽星,周明,孙茂松. 基于层次结构的多策略中文微博情感分析和特征抽取[J]. 中文信息学报,2012(1):73-83.
[9] WILSON T, WIEBE J, HOFFMANN P. Recognizing contextual polarity in phrase-level sentiment analysis[J]. International journal of computer applications, 2005, 7(5):347-354.
[10] TURNEY P D, LITTMAN M L. Measuring praise and criticism:inference of semantic orientation from association[J]. ACM transactions on information systems, 2003, 21(4):315-346.
[11] DUMAIS S T.Latent semantic analysis[J].Annual review of information science and technology,2004,38(1):188-230.
[12] BLEI D M,LAFFERTY J D.A correlated topic model of science[J].The annals of applied statistics,2007,1(1):17-35.
[13] MIKOLOV T, CHEN K, CORRADO G, et al. Efficient estimation of word representations in vector space[J/OL]. Computer science, 2013.[2017-04-30].https://arxiv.org/abs/1301.3781.
[14] MIKOLOV T, SUTSKEVER I, CHEN K, et al. Distributed representations of words and phrases and their compositionality[J/OL]. Advances in neural information processing systems, 2013.[2017-04-30].https:/arxiv.org/abs/1310.4546.
[15] 魏慧玲. 文本情感分析在产品评论中的应用研究[D]. 北京:北京交通大学, 2014.
[16] 杨伟文,刘新. 品牌认知对消费者购买行为的影响[J]. 商业研究,2010(3):158-162.
[17] 龚伟勋,舒敏,陈章旺. 网络口碑传播对顾客品牌认知的影响[J]. 武汉工程大学学报,2010(4):73-77.
