论文部分内容阅读
微博信息推荐本质上属于微博信息组织的内容,由于用户信息的非结构化带来的推荐不准确的问题成为影响微博发展的制约因素。笔者针对用户信息的非结构化问题,从微博信息组织的视角出发,以简化的本体架构——微本体为数据结构来构建微博信息推荐系统的底层数据库,结合改进的LDA模型进行微博信息和用户的推荐,将有效的提高微博信息推荐的精度和质量。实验证明与传统LDA算法相比,基于Folksonmy和Ontology融合的微博信息推荐算法更加可靠。
The recommendation of the microblogging information essentially belongs to the content of the microblogging information organization. The problem of the inaccurate recommendation caused by the unstructured user information has become a constraint factor for the development of the microblog. In view of the unstructured information of user information, from the perspective of Weibo information organization, this paper constructs the underlying database of Weibo information recommendation system with a simplified ontology architecture - Micro Ontology as data structure, and combines the improved LDA model with Weibo Information and user’s recommendation will effectively improve the accuracy and quality of Weibo recommendation. Experiments show that compared with the traditional LDA algorithm, the recommendation algorithm of Weibo information based on Folksonmy and Ontology fusion is more reliable.