摘要网络购物的普及使得商品评论信息获得的关注度不断提升，有关评论信息可信度的研究也成为了当前文本信息领域的研究热点。通过对目前评论信息可信度的相关研究进行深入分析，了解现有评论信息可信度计量研究中存在的问题，最终选择利用输出值可以取任意实数的支持向量回归（SVR）算法完成对评论信息可信度的量化。首先在对商品评论信息可信度影响因素进行实证研究的基础上，从内容完整性、情感平衡性、时效性以及评论发布者身份明确性四个特征文度选取可量化特征；接着构建特征词典并开发可信度特征量化处理软件完成特征的量化；然后，借助机器学习方法，利用LIBSVM工具完成评论信息可信度的计算；最后，通过网格寻优法调整相关参数获取优化的模型，获得评论信息可信度的优化计算结果。在将实验结果与人工标注平均值及评论内容进行匹配分析后，发现该可信度计量模型较好地完成了对商品评论信息可信度的预测及其可信度类型有效识别。19805
关键词 商品评论信息 可信度量化 SVR 机器学习 数据挖掘
毕业论文外文摘要
Title Analysis of Online Product Reviews Credibility Based on
Support Vector Regression model
Abstract
The popularity of online shopping makes product reviews received rising attention, researchers about reviews information credibility has become a hotspot of the current text field. Through in-depth analysis of current research on reviews information credibility , understanding the existing problems in the Research on reviews information credibility measurement and then selected the support vector regression algorithm whose output value can be any real number to quantify the reviews information credibility. Firstly, according to the empirical research of reviews information credibility factors , they selected content integrity, sentimental balance, review timeliness and credit rating of the publisher as four features dimensions that can be quantified. Secondly, build the feature dictionary and develop the credibility characteristic quantification processing software to complete feature quantification ; Thirdly, with the help of machine learning methods, they use LIBSVM tools to complete reviews credibility calculation. Finally, use the Grid Search method to revise the parameters , in order to get the optimize results. After matching analysis the experimental results with manual annotation average and reviews content, it was found that the credibility measurement model forecasts the credibility value and the type of product reviews effectively.
Keywords product reviews credibility measurement SVR machine learning
data mining
目 次
1 绪论 1
1.1 研究背景及意义 1
1.1.1 研究背景 1
1.1.2 研究意义 2
1.2 研究现状 3
1.2.1 商品评论信息的可信度影响因素研究现状 4
1.2.2 商品评论信息可信度计量研究现状 5
1.2.3 现有研究存在的问题 7
1.3 本文的主要研究内容 7
1.4 本研究的特色 8 基于SVR支持向量回归模型的在线商品评论信息可信度分析研究:http://www.751com.cn/shuxue/lunwen_11331.html