商品评论情感分析系统的设计与实现

发布时间：2018-08-12 21:09

【摘要】：在过去的二三十年内,网上购物深刻地影响着我国广大普通消费者的消费行为,同时对于企业来说,在面临新兴的电子商务市场时,应该充分发挥其在生产和销售等环节的指导和决策作用。在电子商务网站上有大量的商品评论数据,这些数据对营销策略具有很重要的指导意义。可以通过分析商品评论文本,确定自己产品优缺点、与竞品对比等。但是并没一款适合的软件满足企业的需求,所以作者所在公司的部门提出建立一个商品评论情感分析系统,为业务部门提供技术支持。本文以手机产品的评论为例进行深入的研究,目标是建立一套完整的商品评论情感分析系统,应用于分析手机产品的优点和缺点。本人参与了系统从分析到应用的全部过程,首先完成原始评论文本的处理,包括分词和词性标注等。其次尝试多种文本分类方法,进行优选,确定并实现适合于手机产品使用的分类方法,并建立大量文本分类规则库。然后探索不同的文本情感分析方法,比较这些方法的优缺点,选出最合适的方法并实现,最后完成商品评论情感分析系统。在系统的实现过程中建立了文本分类库和文本规则库,并通过训练数据,建立了贝叶斯情感分类模型。在商品评论情感分析系统完成后,根据客户的反馈意见,不断进行测试修改,在此过程中积累了大量文本分析的经验,逐步提高了分析结果的准确度,并将该方法应用到产品销售的各个方面。已经为多款产品提供了完整全面的商品评论分析报告,实现了商品评论情感分析的目的。
[Abstract]:In the past two or three decades, online shopping has profoundly affected the consumer behavior of the vast majority of ordinary consumers in our country. At the same time, for enterprises, when facing the emerging e-commerce market, Should give full play to its production and sales and other links of guidance and decision-making role. There are a lot of commodity review data on e-commerce website, these data have very important guiding significance to marketing strategy. Through the analysis of commodity review text, determine their own product strengths and weaknesses, and competitive products and so on. But there is not a suitable software to meet the needs of the enterprise, so the department of the company proposed to establish a commodity review emotional analysis system to provide technical support for the business department. This paper takes the mobile phone product comment as an example to carry on the thorough research, the goal is to establish a set of complete commodity comment emotion analysis system, applies to analyze the mobile phone product the merit and the shortcoming. I participated in the whole process from analysis to application. First, I completed the processing of the original comment text, including participle and part of speech tagging. Secondly, we try a variety of text classification methods, select the best, determine and implement the classification methods suitable for mobile phone products, and establish a large number of text classification rules database. Then explore different text emotional analysis methods, compare the advantages and disadvantages of these methods, select the most appropriate method and implement, and finally complete the commodity review emotional analysis system. The text classification database and text rule database are established in the process of implementation of the system, and Bayesian emotional classification model is established by training data. After the completion of the commodity review emotional analysis system, according to the feedback of the customer, the testing and modification are carried out continuously. In the process, a large number of experience of text analysis has been accumulated, and the accuracy of the analysis results has been improved step by step. The method is applied to all aspects of product sales. Has provided a complete and comprehensive commodity review analysis report for a number of products, achieving the purpose of commodity review emotional analysis.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.1

【参考文献】