基于地理层次优化的排序推荐算法
首发时间:2020-04-07
摘要:近年来随着基于地理位置的服务类应用的日益普及,兴趣点推荐算法的研究也取得了一定的进展。但是当前多数研究工作并未充分利用兴趣点特有的地理位置信息来加速模型收敛过程并提高模型推荐准确性,为了解决以上问题,本文首先提出了一种基于地理层次的采样方法来生成偏序对,接着提出具有地理感知的小批量梯度下降算法,最后通过直接优化贝叶斯排序指标使得提供给用户的推荐列表具有更好的排序质量。
For information in English, please click here
Ranking recommendation algorithm based on geographic hierarchy optimization
Abstract:In recent years, with the increasing popularity of geographic-based service applications, research on Point-of-Interest recommendation algorithms has also made some progress. However, most of the current research work does not make full use of the geographic location information unique to the POI to accelerate the model convergence process and improve the accuracy of the model recommendation. In order to solve the above problems, this paper first proposes a sampling method based on geographic hierarchy to generate partial order pairs, Then, a geographically-aware mini-batch gradient descent algorithm is proposed. Finally, by directly optimizing the Bayesian ranking index, the recommendation list provided to users has better ranking quality.
Keywords: computer application ranking recommendation geographical hierarchy optimization
基金:
引用

No.****
动态公开评议
共计0人参与
勘误表
基于地理层次优化的排序推荐算法
评论
全部评论