基于文本挖掘的用户画像研究.pdf

汕头大学硕士学位论文基于文本挖掘的用户画像研究I摘要随着互联网的快速发展与日益普及,网络营销的价值也逐渐得到重视与认可,因此电子商务得到迅猛发展,而电子商务网站也日益成为大多数人生活中的“必需品”。而且越来越多的商家也希望借助电子商务在互联网销售市场的激烈竞争中立于不败之地。电子商务的不断发展,用户行为的研究是影响其生存发展的重要因素,用户行为研究的好坏成了决定消费者去留的关键原因。随着互联网用户规模及电子商务市场的不断扩大,电子商务市场的竞争愈发激烈,把握企业的竞争优势,同时加强企业战略性分化发展的能力,保证企业的可持续性发展是所有电子商务企业的共同目标。电子商务企业为了优化在网站上的营销活动和网站运营开销,都要投入大量的资源进行网站用户行为分析。随着产业的迅速发展,电子商务企业需要采用更先进的手段对网站进行网站用户行为分析,并构建自己的用户画像,并且大多电子商务网站已经积累了足够多的用户消费行为信息来进行用户细分等相关分析。因此,本文通过对我国主流电子商务网站的数据进行相关采集与研究,并提出了用户画像构建的研究策略,并提出将用户属性分为基础属性标签,行为属性标签,价值属性标签,社交属性标签。并采用概率与信息熵的方法对用户数据进行分词,采用层次分析法对用户价值属性进行分析,得到用户的价值属性标签,并通过定义相关的规则,以此来构建用户画像,并采用k-means对构建的用户画像进行聚类。关键词:电子商务;用户画像;用户研究;用户聚类分析;k-means万方数据汕头大学硕士学位论文基于文本挖掘的用户画像研究IIAbstractWith the rapid development of , the value work marketing has gradually received attention and recognition. merce has been developed rapidly ,and the merce website has e a necessity for most people in life. More and more businesses also hope to stand firm and invincible in the fierce competition by means of merce on the sales market .With the continuous development of merce, the user behavior has e more and more important for its development, user behavior research has e the key reasons for retaining consumers .Along with the user scale and the merce market continues to expand, the merce petition is increasingly intense, grasp petitive advantage of enterprises, and strengthen the ability of enterprise strategic development, guarantee the sustainable development of enterprises is mon goal of allenterprise business enterprise have to invest a lot of resources website user behavior analysis in order to optimize the website marketing and website operation cost. With the rapid development of industry, merce enterprises need to use more advanced methods for website user behavior analysis, and construct the user portrait of himself, and most of the electronic commerce website has accumulated enough consumer behavior information to do user segment