摘要构建用户模型,给用户提供个性化服务已成为互联网下的必然趋势。在互联网的基础 上,网络社交已成为常态,常见的社交平台有微博、微信、豆瓣网等,其中新浪微博是中 国最大的社交媒体,是人们日常生活的重要组成部分。用户在社交平台中浏览、发布信息 等,在一定程度上反映了用户的关注热点,且目前较多社交平台提供标签标注功能,用户 不仅可以为资源标注标签,也可以为自己选择和定义标签,这些用户标签体现了用户的自 身属性及兴趣所在,对用户建模与用户兴趣发现具有重要意义。本文以新浪微博为平台, 以用户微博内容和用户标签作为数据源,构建用户模型。本文构建的用户模型由两部分组 成:其一为基于微博内容的用户兴趣表示,即:首先对用户微博内容进行中文分词、特征 抽取等预处理,然后采用 TF-IDF 方法进行特征权重计算,接着将用户的微博内容表示为 向量空间模型;其二为基于用户标签的用户兴趣表示,即:通过文档频次法对用户标签进 行特征抽取后采用标签频次法进行特征权值计算,接着也将用户标签表示成向量空间模型; 最终,本文将以上两个兴趣表示合成,得到用户的兴趣模型。79504
毕业论文关键词 用户建模 特征提取 用户生成内容 用户标签 向量空间模型
毕 业 论 文 外 文 摘 要
Title User Modeling on Social Networks——Using User Tags and Weibo Content for User Modeling
Abstract It is the trend to user modeling and provide personalized service in our days。 And with social networking becoming popular, social platforms like Weibo, WeChat, Douban are very common,
among then Weibo is China's largest social media and plays an important role in people's daily life。 Users browsed and issued Weibo content in the social platform reflects the user’s concerns。 And now more and more social platforms provide label function, users can not only label for the resource, but also label for yourself。 These user tags reflecting user's attributes and interests, it is of great significance for user modeling and user interest mining。 This paper collects user’s
weibo content and tags as the data resource for user modeling。 The user model consists of two parts, one, user interest representation about weibo content: first , pretreatment like Chinese word segmentation, feature extraction, and then use TF-IDF method to compute the characteristic value, after that, the user's weibo content expressed by vector space model (VSM);Another part, user interest representation about the user tags: document frequency
statistics method for feature extraction and then through the word frequency method for computing the characteristic value and the user tags also expressed by VSM。 Finally, combining two parts mentioned above is the end user model。
Keywords User modeling; Feature extraction; User generated content; User tag; Vector space model
目 次
1 绪论 3
1.1 选题背景 3
1.2 研究意义 3
1.3 本文的研究内容及章节安排