首页> 外文会议>International conference on web information systems engineering >Geographical Constraint and Temporal Similarity Modeling for Point-of-Interest Recommendation
【24h】

Geographical Constraint and Temporal Similarity Modeling for Point-of-Interest Recommendation

机译:地理约束和时间相似性建模对兴趣点推荐

获取原文

摘要

People often share their visited Points-of-Interest (Pols) by "check-ins". On the one hand, human mobility varies with each individual but still implies regularity. Check-ins of an individual tend to localize in a specific geographical range. We propose a novel model to capture personalized geographical constraint of each individual. On the other hand, Pols reflect requirements of people from different aspects. Usually, places of different functions show different temporal visiting distributions and places of similar function share similar visiting pattern in temporal aspect. Temporal distribution similarity can be used to characterize functional similarity. Based on the findings above, this paper introduces improved collaborative filtering models by jointly taking advantages of geographical constraint and temporal similarity. Experimental results on real data collected from Gowalla and JiePang demonstrate the effectiveness of our models.
机译:人们经常通过“登记”分享他们访问的兴趣点(POL)。一方面,人类流动性因每个人而异,但仍然是规律性。个人的核心核心倾向于定位特定的地理范围。我们提出了一种小说模型来捕获每个人的个性化地理限制。另一方面,POLS反映了来自不同方面的人的要求。通常,不同函数的位置显示不同的时间访问分布以及类似功能的地方在时间方面共享类似的访问模式。时间分布相似度可用于表征功能相似性。基于上面的研究结果,本文通过共同采取地理约束和时间相似性来介绍改进的协作滤波模型。 Gowalla和Jiepang收集的实验结果展示了模型的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号