基于在线评论情感分析的手机推荐方法
|
摘要
手机的普及为生活带来便利,但由于品牌多、定位不同等导致消费者难以快速准确的选择合适的机型。为解决这个问题,提出了一种基于在线评论情感分析的手机推荐方法,用于帮助消费者寻找符合购买意愿的手机。首先,集成基础情感词典,构建程度副词、否定词、手机领域评论用语等相关词典,再根据消费者需求对手机各项参数进行多属性组合,并计算每条手机评论的情感值,将评论文本的情感模糊值转换为准确值,最后结合手机属性和评论的情感值定义特征矩阵,进而使用余弦相似度的方法得到top-k款与消费者需求相似的手机推荐给用户。对比传统的使用单一情感词典的推荐方法,结果表明该推荐方法能够有效的提高推荐结果的采纳度。
The popularity of mobile phones has brought convenience to life. But due to various brands and different positioning, it is difficult for consumers to quickly and accurately select the right model. In order to solve this problem, this paper proposes a recommendation method for mobile phone based on sentiment analysis of online reviews, which helps consumers find a mobile phone that meets their purchase intention. First, the basic sentiment dictionary is integrated to build related dictionaries such as degree adverbs, negative words and reviews terms in the field of mobile phone. Then, various parameters of the mobile phone are combined with multiple attributes according to the demand of consumer, and the sentiment value of each mobile phone reviews is calculated. The sentiment fuzzy value of the reviews text is converted into the accurate value. Finally, combine the mobile phone attributes and the sentiment value of the reviews to define the characteristic matrix, and then by using the method of cosine similarity get the top-k types mobile phone which is similar to the consumer's demand. The results show that this method can effectively help recommend the consumers purchasing mobile phones.
The popularity of mobile phones has brought convenience to life. But due to various brands and different positioning, it is difficult for consumers to quickly and accurately select the right model. In order to solve this problem, this paper proposes a recommendation method for mobile phone based on sentiment analysis of online reviews, which helps consumers find a mobile phone that meets their purchase intention. First, the basic sentiment dictionary is integrated to build related dictionaries such as degree adverbs, negative words and reviews terms in the field of mobile phone. Then, various parameters of the mobile phone are combined with multiple attributes according to the demand of consumer, and the sentiment value of each mobile phone reviews is calculated. The sentiment fuzzy value of the reviews text is converted into the accurate value. Finally, combine the mobile phone attributes and the sentiment value of the reviews to define the characteristic matrix, and then by using the method of cosine similarity get the top-k types mobile phone which is similar to the consumer's demand. The results show that this method can effectively help recommend the consumers purchasing mobile phones.
引文
[1] 冯登国,张敏,李昊.大数据安全与隐私保护[J].计算机学报,2014,37(1):246-258.
[2] 钱宇华,成红红,梁新彦,等.大数据关联关系度量研究综述[J].数据采集与处理,2015,30(6):1 147-1 159.
[3] 田依林,滕广青,黄微.社会化在线评论情感因素测量与有用性分析[J].现代情报,2018(6):19-27.
[4] 周国强,刘旭,杨锡慧.基于情感权重的用户协同推荐模型[J].小型微型计算机系统,2016,37(5):938-942.
[5] 任巨伟,杨亮,吴晓芳,等.基于情感常识的微博事件公众情感趋势预测[J].中文信息学报,2017,31(2):169-178.
[6] 陈磊,李俊.基于词向量的文本特征选择方法研究[J].小型微型计算机系统,2018,39(5):991-994.
[7] 刘洪超,黄居仁,侯仁魁,等.基于语言学特征向量和词嵌入向量的汉语动词事件类型预测[J].中文信息学报,2018,32(1):26-33.
[8] 田星,郑瑾,张祖平.基于词向量的Jaccard相似度算法[J].计算机科学,2018,45(7):186-189.
[9] 郭正斌,张仰森,蒋玉茹.一种面向文本分类的特征向量优化方法[J].计算机应用研究,2017,34(8):2 299-2 302.
[10] 赵妍妍,秦兵,石秋慧,等.大规模情感词典的构建及其在情感分类中的应用[J].中文信息学报,2017,31(2):187-193.
[11] 蒋盛益,黄卫坚,蔡茂丽,等.面向微博的社会情绪词典构建及情绪分析方法研究[J].中文信息学报,2015,29(6):166-171,202.
[12] 董红斌,滕旭阳,杨雪.一种基于关联信息熵度量的特征选择方法[J].计算机研究与发展,2016,53(8):1 684-1 695.
[13] 哈金才.数据挖掘算法的评价标准与方法[J].微电子学与计算机,2006,23(12):195-196.
[2] 钱宇华,成红红,梁新彦,等.大数据关联关系度量研究综述[J].数据采集与处理,2015,30(6):1 147-1 159.
[3] 田依林,滕广青,黄微.社会化在线评论情感因素测量与有用性分析[J].现代情报,2018(6):19-27.
[4] 周国强,刘旭,杨锡慧.基于情感权重的用户协同推荐模型[J].小型微型计算机系统,2016,37(5):938-942.
[5] 任巨伟,杨亮,吴晓芳,等.基于情感常识的微博事件公众情感趋势预测[J].中文信息学报,2017,31(2):169-178.
[6] 陈磊,李俊.基于词向量的文本特征选择方法研究[J].小型微型计算机系统,2018,39(5):991-994.
[7] 刘洪超,黄居仁,侯仁魁,等.基于语言学特征向量和词嵌入向量的汉语动词事件类型预测[J].中文信息学报,2018,32(1):26-33.
[8] 田星,郑瑾,张祖平.基于词向量的Jaccard相似度算法[J].计算机科学,2018,45(7):186-189.
[9] 郭正斌,张仰森,蒋玉茹.一种面向文本分类的特征向量优化方法[J].计算机应用研究,2017,34(8):2 299-2 302.
[10] 赵妍妍,秦兵,石秋慧,等.大规模情感词典的构建及其在情感分类中的应用[J].中文信息学报,2017,31(2):187-193.
[11] 蒋盛益,黄卫坚,蔡茂丽,等.面向微博的社会情绪词典构建及情绪分析方法研究[J].中文信息学报,2015,29(6):166-171,202.
[12] 董红斌,滕旭阳,杨雪.一种基于关联信息熵度量的特征选择方法[J].计算机研究与发展,2016,53(8):1 684-1 695.
[13] 哈金才.数据挖掘算法的评价标准与方法[J].微电子学与计算机,2006,23(12):195-196.