论文部分内容阅读
高速发展的微博带来信息富余,也带来了信息过载,不断新增的非结构化微博文本内容和复杂的社会网络关系导致个性化推荐难以实施.针对微博网站特征,提出一种基于信息传播模拟的协同过滤推荐模型并给出推荐框架图,解决推荐的数据稀疏性和冷启动问题.首先,通过自然语言处理技术处理非结构化文本内容,获取关键词为推荐对象,构建用户-关键词偏好模型;然后,采用一阶马尔可夫随机游走模拟用户偏好在社会网络中的传播过程,得到用户-关键词偏好矩阵.实验使用来自新浪微博的数据集,采用平均绝对误差、准确率和召回率三个指标评价推荐模型,并与基准模型进行对比.实验结果表明,因整合了社会网络结构信息,基于信息传播的协同过滤推荐模型的效率比基准模型有明显提高.
The rapid development of microblogging brings information redundancy, but also brought information overload, and constantly add unstructured microblogging text content and complex social network relationships lead to personalized recommendations difficult to implement.According to the characteristics of Weibo site, a A collaborative filtering recommendation model based on information dissemination simulation is proposed and a recommended framework is proposed to solve the proposed data sparsity and cold start problem.Firstly, the unstructured text content is processed by natural language processing technology, the key words are recommended objects, - Keyword preference model, and then use the first-order Markov random walk to simulate the process of user preferences in social networks and get the user-keyword preference matrix.Experiments use data sets from Sina Weibo with average absolute error , Accuracy and recall, and compared with the benchmark model.Experimental results show that the efficiency of collaborative filtering recommendation model based on information dissemination is significantly improved compared with the benchmark model because of the integration of social network structure information.