基于机器学习的指标协同自适应推荐算法

Adaptive recommendation algorithm for index collaboration based on machine learning

为解决电力通信指标规划复杂且难以精确预测用户所需内容等问题,通过度量用户的选择频次、重要专家评分以及指标系统的使用时长,将语义内容融入到协同过滤中,并采用机器学习进行指标衍生,实现基于机器学习指标衍生的自适应关键内容推荐。仿真表明,采用机器学习的自适应推荐算法能够有效预测用户所需内容,算法推荐准确性可以达到90%以上。

To address the current problems of complex planning of power communication index and difficulty in accurately predicting the contents required by users,the semantic information is incorporated into the collaborative filtering by measuring the selection frequency of users,important expert ratings and the usage duration of the indicator system.And the machine learning is used for indicator derivation to achieve adaptive critical content recommendation based on machine learning indicator derivation.The simulation shows that the adaptive recommendation algorithm based on machine learning can effectively predict the content required by users,and the accuracy of the algorithm recommendation can reach more than 90%.