【摘要】 为了准确挖掘用户偏好，提升数据挖掘效率，避免信息过载问题，提出基于虚拟最近邻矩阵的用户偏好数据挖掘方法。在考虑用户特征与用户偏好的条件下，采用协同过滤算法计算用户相似性，通过目标用户和邻近用户的相似性对目标项目展开预测评分，并按照从高到低的顺序对评分结果进行排序，构建虚拟最近邻矩阵，结合评分结果，获取用户偏好信息。以用户偏好预测结果为依据，通过K-means聚类与关联规则算法组建用户偏好挖掘模型，获取用户偏好数据。仿真结果表明，所提方法可以以较高的准确性和效率完成用户偏好数据挖掘，数据挖掘结果较为理想。更多还原

【Abstract】 In order to accurately mine user preference, improve data mining efficiency, and avoid information overload, this article puts forward a method of mining user preference data based on virtual nearest neighbor matrix. With consideration of user features and user preferences, the collaborative filtering algorithm was adopted to calculate the user similarity. And then, the target item was predicted and scored through the similarity between target user and adjacent users. Meanwhile, the scoring results were sorted from high to low. Moreover, a virtual nearest neighbor matrix was constructed. Based on the scoring results, user preference information can be obtained. On the basis of the prediction result, K-means clustering and association rule algorithm were adopted to build a user preference mining model. Finally, user preference data were obtained. Simulation results show that the proposed method can mine the user preference data with high accuracy and efficiency, and the data mining results are ideal.更多还原