论文部分内容阅读
[目的/意义]针对某些包含多级用户和多级资源的异质网络,在总结其各种异质模态的基础上提出一种多维度推荐框架MDRM,向目标用户进行好友和资源的推荐。[方法/过程]通过对用户、资源划分等级,在分析各种网络模态的基础上建立其异质关系;利用情感倾向分析得到二级用户——二级资源评分矩阵,借助协同过滤算法,实现同级用户和二级资源的推荐;基于异质关系,实现一级用户和一级资源的推荐,最终实现多维度推荐。[结果/结论]在以豆瓣网数据作为数据集的实验中取得了较好的效果,说明MDRM模型适合某些异质网络资源的推荐。
[Purpose / Significance] Aiming at some heterogeneous networks with multi-level users and multi-level resources, a multidimensional recommendation framework (MDRM) is proposed based on the summarization of various heterogeneous modalities. It aims to make friends and resources to target users recommend. [Methods / Processes] Through the analysis of various network modalities, users and resources are divided into different levels based on their heterogeneity. Second-level users’ secondary resource rating matrixes are obtained through the analysis of affective tendencies. With the collaborative filtering algorithm, To achieve the recommendation of peer users and secondary resources; to realize the recommendation of first-level users and first-level resources based on the heterogeneous relationship, and finally achieve multi-dimensional recommendation. [Results / Conclusions] Good results have been obtained in experiments using Douban data as a data set, indicating that the MDRM model is suitable for the recommendation of some heterogeneous network resources.