个性化信息标志着用户从被动的信息消费者变为主动的信息创造者,而个性化信息推荐是通过用户以往的标记行为,向用户提供将来可能用到的信息。这意味着向每一个用户推荐一张个性化的信息清单,这个清单既取决于资源的内容也取决于用户本身。针对已有的基于张量分解中的塔克分解的信息推荐的缺点——三维立方核心张量导致的个性化信息推荐的立方运行时间加大,尤其是大的数据集,提出新的张量分解方法来缩短系统运行时间,最后进行方法的优化验证。 Personalized information indicates that the user changes from a passive information consumer to an active information creator, and the personalized information recommendation provides the user with information that may be used in the future through the user’s past marking behavior. This means recommending to each user a personalized list of messages that depends on the content of the resource and on the user itself. Recommended Defects for Tuck Decomposition Based on Existing Tensor Decompositions - Personalized Information Caused by Three-Dimensional Cortical Tensor The recommended cubic runtime increases, especially for large data sets, proposing new tensors Decomposition method to shorten the system running time, and finally the method of optimization and verification.