大数据知识服务的情景化推荐,就是将用户情景信息引入个性化推荐过程,结合云计算技术提出的一种大数据知识服务方法。该方法首先计算大数据知识服务用户的情景相似度,并构造与目标用户当前情景相似的情景集合,建立基于项目评分情景的评分矩阵,最后进行云计算Mapreduce化的情景化推荐。实验表明,该方法获得了比传统推荐方法更低的MAE误差值,可成功应用于面向大数据知识服务的情景化推荐。
The contextual recommendation for the big data knowledge service is a method of the big data knowledge service that inputs the contextual information of the user into the personalized recommendation process and combines with the cloud computing technology. This method firstly calculates the contextual similarity of users of the big data knowledge service, and builds the contextual set that is similar to the current scene of the target user, establishes the score matrix based on the project score scene, and finally runs the cloud computing Mapreduce contextual recommendation. Experimental results show that this method achieved lower error values of MAE than the traditional recommendation method, and could apply in the contextual recommendation oriented the big data knowledge service.