Review of Point of Interest Recommendation Systems in Location-Based Social Networks

doi:10.3778/j.issn.1673-9418.2112037

Abstract

Abstract:

Point of interest recommendation is recently one of the hotspots in the field of location-based social networks and recommendation systems. Understanding the research status of the point of interest recommendation in location-based social networks can provide a direction for the next step of work. The recent literatures of the point of interest recommendation systems are analyzed. Firstly, the definition is introduced, and the difference from traditional recommendation is discussed from three aspects: influencing factors, recommendation approaches and existing problems. Secondly, the general framework of the point of interest recommendation is proposed, which includes data sources, recommendation approaches and evaluation. Based on this framework, the various influencing factors are introduced, the current recommendation algorithms are generalized, and the evaluation metrics are summarized. Meanwhile, the representative works are analyzed, the research contents and characteristics of each type of methods are summarized in detail, and their advantages and limitations are evaluated. Finally, the challenges and potential directions for possible extensions in this filed are summarized and prospected, and the future research trends and development directions are concluded.

Key words: location-based social networks, recommendation systems, point of interest recommendation, influencing factor

摘要：

兴趣点推荐是近年来位置社交网络和推荐系统领域研究的热点之一,了解兴趣点推荐在位置社交网络方面的发展现状,有利于为下一步的研究提供方向。对国内外兴趣点推荐系统的相关文献进行梳理,首先介绍了兴趣点推荐系统的概念,并从影响推荐的因素、推荐方法和推荐存在的问题三方面探讨其与传统推荐的区别。然后提出了兴趣点推荐系统的基本框架,该框架包含了数据来源、推荐方法和算法评价三个核心部分。以该框架为基础,介绍了影响兴趣点推荐的多种因素,归纳了现有的兴趣点推荐算法,总结了算法的评价指标。同时对代表性工作进行了分析介绍,详细总结了各种方法的研究内容与特点,并评价了其优势与不足。最后对该领域所面临的挑战和潜在的研究方向进行了总结与展望,给出了未来的研究趋势和发展方向。

关键词: 位置社交网络, 推荐系统, 兴趣点推荐, 影响因素

CLC Number:

TP391

CHEN Jiangmei, ZHANG Wende. Review of Point of Interest Recommendation Systems in Location-Based Social Networks[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(7): 1462-1478.

陈江美, 张文德. 基于位置社交网络的兴趣点推荐系统研究综述[J]. 计算机科学与探索, 2022, 16(7): 1462-1478.

Figures/Tables 11

Fig.1 Location-based social network

Fig.2 Users interaction data

Fig.3 Category of recommendation algorithms

Table 1 Difference between traditional recommendation and POI recommendation

比较类型	兴趣点推荐系统	传统的推荐系统
位置维度	有	无
信息来源	签到数据、情景信息	评分信息
数据类型	隐式数据	显式数据
推荐结果展示	基于评价的方式、列表方式	基于评价的方式
解决问题	数据稀疏、冷启动、序列推荐、动态推荐、个性化推荐、异地推荐	数据稀疏、冷启动

Fig.4 Framework of POI recommendation system

Table 2 Comparison of typical algorithms in influencing factors of POI recommendation

代表算法	影响推荐因素	说明	优点	局限性	代码链接
USG^[7]	用户偏好、地理、社交	将用户偏好融合到具有社会和地理影响的推荐框架	整合了多种情景信息	无法解决数据稀疏问题	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/ USG.zip
GeoMF^[28]	地理	融合地理信息到加权矩阵分解算法模型	捕捉了用户的空间聚类现象	只考虑了地理位置影响	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/GeoMF.zip
iGSLR^[15]	地理、社交	将社会影响和个性化地理影响整合到统一的推荐框架	实现了个性化的推荐	无法解决冷启动问题和数据稀疏问题	https://github.com/camcochet/iGSLR-Personalized-Geo-Social-Location-Recommendation
LRT^[14]	时间	提出时间的连续性和差异性特征,融入矩阵分解框架中	模拟了时间特征影响;实现了动态推荐	推荐准确率较低;只考虑了时间影响	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/LRT.zip
GTSCP^[17]	地理、社交、时间、内容、流行度	将所有情景信息融入联合概率生成模型	适用于异地推荐场景;缓解数据稀疏问题	未解决兴趣点特征缺失的不足	N/A
LGLMF^[29]	地理、流行度	考虑用户的主要活动区域及该区域内每个位置的相关性	可缓解数据稀疏性问题	未充分考虑其他情景信息的影响	https://paperswithcode.com/paper/lglmf-local-geographical-based-logistic#code
SLGMF^[26]	地理、社交	融合社交和局部地理因素影响	可缓解数据稀疏性问题	融合的情景信息较有限	N/A
CRQA^[47]	地理、内容	将空间和地理信息融合到联合推理模型中	可扩展到其他领域的推荐任务	只考虑了空间和文本推理	https://paperswithcode.com/paper/joint-spatio-textual-reasoning-for-answering#code
MM-Gated-XAtt^[48]	内容、图像	从各模态中提取相关信息来获取文本和图像间的交互	有效捕获多模态间的交互	未充分利用兴趣点与用户信息	https://github.com/danaesavi/poi-type-prediction

Table 2 Comparison of typical algorithms in influencing factors of POI recommendation

代表算法	影响推荐因素	说明	优点	局限性	代码链接
USG^[7]	用户偏好、地理、社交	将用户偏好融合到具有社会和地理影响的推荐框架	整合了多种情景信息	无法解决数据稀疏问题	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/ USG.zip
GeoMF^[28]	地理	融合地理信息到加权矩阵分解算法模型	捕捉了用户的空间聚类现象	只考虑了地理位置影响	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/GeoMF.zip
iGSLR^[15]	地理、社交	将社会影响和个性化地理影响整合到统一的推荐框架	实现了个性化的推荐	无法解决冷启动问题和数据稀疏问题	https://github.com/camcochet/iGSLR-Personalized-Geo-Social-Location-Recommendation
LRT^[14]	时间	提出时间的连续性和差异性特征,融入矩阵分解框架中	模拟了时间特征影响;实现了动态推荐	推荐准确率较低;只考虑了时间影响	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/LRT.zip
GTSCP^[17]	地理、社交、时间、内容、流行度	将所有情景信息融入联合概率生成模型	适用于异地推荐场景;缓解数据稀疏问题	未解决兴趣点特征缺失的不足	N/A
LGLMF^[29]	地理、流行度	考虑用户的主要活动区域及该区域内每个位置的相关性	可缓解数据稀疏性问题	未充分考虑其他情景信息的影响	https://paperswithcode.com/paper/lglmf-local-geographical-based-logistic#code
SLGMF^[26]	地理、社交	融合社交和局部地理因素影响	可缓解数据稀疏性问题	融合的情景信息较有限	N/A
CRQA^[47]	地理、内容	将空间和地理信息融合到联合推理模型中	可扩展到其他领域的推荐任务	只考虑了空间和文本推理	https://paperswithcode.com/paper/joint-spatio-textual-reasoning-for-answering#code
MM-Gated-XAtt^[48]	内容、图像	从各模态中提取相关信息来获取文本和图像间的交互	有效捕获多模态间的交互	未充分利用兴趣点与用户信息	https://github.com/danaesavi/poi-type-prediction

Fig.5 Bipartite graph network

Fig.6 Structure of convolutional neural network

Table 3 Comparison of typical algorithms in POI recommendation methods

推荐算法类别	代表算法	说明	优点	局限性	代码链接
基于矩阵分解的兴趣点推荐	ASMF^[56]	考虑三类朋友来寻找每个用户的潜在兴趣点,并将地理和分类信息融入矩阵分解框架中	考虑了朋友的影响,同时可缓解数据稀疏问题	融合的情景信息较有限	https://paperswithcode.com/paper/point-of-interest-recommendations-learning#code
	GeoMF^[28]	利用加权矩阵分解算法融合地理信息	捕捉了用户的空间聚类现象	考虑的情景信息单一	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/GeoMF.zip
	STACP^[57]	将时空信息合并到矩阵分解模型中	可缓解数据稀疏问题;实现动态推荐	无法解决冷启动问题	https://paperswithcode.com/paper/joint-geographical-and-temporal-modeling#code
基于图嵌入的兴趣点推荐	UP2VEC^[35]	利用异构图融合用户的社交、地理和时间信息	解决社交网络的异构性问题	模型的复杂度较高	N/A
	RELINE^[60]	利用基于图的方法从各信息关系图中学习用户和兴趣点表示,并嵌入到潜在空间中	解决冷启动问题	未考虑节点属性	https://paperswithcode.com/paper/reline-point-of-interest-recommendations#code
	GLR^[58]	利用基于图嵌入的方法捕获时间序列和用户时间偏好信息	可有效模拟用户复杂的移动行为	可解释性较低	N/A
基于深度学习的兴趣点推荐	DAN-SNR^[70]	利用自注意力机制建模序列信息和社交影响	可有效表征用户与兴趣点的特征	模型的复杂度较高	https://paperswithcode.com/paper/dan-snr-a-deep-attentive-network-for-social#code
	GNN-POI^[68]	利用图神经网络构建用户-兴趣点节点的网络图	适用于各类推荐任务	模型的复杂度较高	N/A
	STAN^[71]	一种时空双向注意模型,充分考虑了相关位置的时空效应	考虑了非相邻位置和非连续签到位置的相关性	融合的情景信息较有限	https://paperswithcode.com/paper/stan-spatio-temporal-attention-network-for-1#code

Table 3 Comparison of typical algorithms in POI recommendation methods

推荐算法类别	代表算法	说明	优点	局限性	代码链接
基于矩阵分解的兴趣点推荐	ASMF^[56]	考虑三类朋友来寻找每个用户的潜在兴趣点,并将地理和分类信息融入矩阵分解框架中	考虑了朋友的影响,同时可缓解数据稀疏问题	融合的情景信息较有限	https://paperswithcode.com/paper/point-of-interest-recommendations-learning#code
	GeoMF^[28]	利用加权矩阵分解算法融合地理信息	捕捉了用户的空间聚类现象	考虑的情景信息单一	http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/code/GeoMF.zip
	STACP^[57]	将时空信息合并到矩阵分解模型中	可缓解数据稀疏问题;实现动态推荐	无法解决冷启动问题	https://paperswithcode.com/paper/joint-geographical-and-temporal-modeling#code
基于图嵌入的兴趣点推荐	UP2VEC^[35]	利用异构图融合用户的社交、地理和时间信息	解决社交网络的异构性问题	模型的复杂度较高	N/A
	RELINE^[60]	利用基于图的方法从各信息关系图中学习用户和兴趣点表示,并嵌入到潜在空间中	解决冷启动问题	未考虑节点属性	https://paperswithcode.com/paper/reline-point-of-interest-recommendations#code
	GLR^[58]	利用基于图嵌入的方法捕获时间序列和用户时间偏好信息	可有效模拟用户复杂的移动行为	可解释性较低	N/A
基于深度学习的兴趣点推荐	DAN-SNR^[70]	利用自注意力机制建模序列信息和社交影响	可有效表征用户与兴趣点的特征	模型的复杂度较高	https://paperswithcode.com/paper/dan-snr-a-deep-attentive-network-for-social#code
	GNN-POI^[68]	利用图神经网络构建用户-兴趣点节点的网络图	适用于各类推荐任务	模型的复杂度较高	N/A
	STAN^[71]	一种时空双向注意模型,充分考虑了相关位置的时空效应	考虑了非相邻位置和非连续签到位置的相关性	融合的情景信息较有限	https://paperswithcode.com/paper/stan-spatio-temporal-attention-network-for-1#code

Table 4 Analysis of research status in POI recommendation

影响因素和问题		推荐策略	推荐算法
影响因素	地理信息	考虑用户签到兴趣点的概率与物理距离相关,提出幂律分布模型（Ye等^[7],2011）	Hybrid
		考虑用户在多个中心点范围内签到,提出多中心的高斯分布模型（Cheng等^[24],2012）	MF
		提出核密度估计方法建模用户的地理偏好（Zhang等^[15],2013）	Hybrid
		提出固定带宽核密度估计方法构建地理偏好模型（Zhang等^[16],2015）	—
		利用自适应核密度估计方法捕捉用户的地理偏好（任星怡等^[20],2017;Si等^[51],2019）	MF等
	用户偏好	将情景信息融入矩阵分解框架（任星怡等^[20],2017;彭宏伟等^[21],2019）	MF
		采用基于用户的协同过滤方法建模签到信息和地理信息（Song等^[18],2019）	CF
		将好友重要性与签到相关性融入协同过滤方法（Zhou等^[19],2019）	CF
		利用卷积神经网络挖掘用户评论中的语义信息（Xing等^[23],2019）	CNN
	社交	利用概率矩阵分解模型建模社交关系（Qian等^[30],2014）	MF
		提出基于朋友的协同过滤方法（彭宏伟等^[21],2019）	CF
		将改进后的直接信任与间接信任融入矩阵分解算法（Xu等^[33],2021）	MF
		利用图嵌入的方法表示用户间的社交关系（Zhu等^[34],2019;Chen等^[36],2021）	GE
		利用图神经网络构建社交网络图（Zhang等^[68],2021）	GNN
	时间	提出基于时间感知的兴趣点推荐模型（Yuan等^[8],2013）	Hybrid
		提出时间的差异性和连续性特征（Gao等^[14],2013）	MF
		提出基于非对称投影的时间感知嵌入方法（Ying等^[37],2019）	—
		利用时间信息捕捉用户的动态偏好（Ma等^[38],2020;Yin等^[39],2021）	RNN、Hybrid
		利用时间序列数据实现下一个兴趣点推荐（Chen等^[64],2020;Huang等^[66],2021）	RNN、Hybrid
	流行度	考虑兴趣点流行度影响用户的决策行为（Yuan等^[8],2013;任星怡等^[20],2017）	Hybrid、MF
		利用文本和图像信息预测兴趣点流行度（Yang等^[49],2019）	—
		将流行度特征结合二维的核密度估计与一维的幂律分布（Si等^[51],2019）	—
	内容	将兴趣点的文本内容嵌入深度学习模型（Chen等^[44],2020）	Hybrid
	内容	采用加权矩阵分解算法融合图像内容和地理信息（Zhang等^[46],2019）	MF
推荐问题	数据稀疏性	利用矩阵分解技术缓解稀疏问题（Gao等^[14],2013;温彦等^[25],2019）	MF
		利用图嵌入表示方法缓解稀疏性（Zhu等^[34],2019）	GE
		采用混合图卷积网络应对稀疏性问题（Zhong等^[67],2020）	GNN
	冷启动	利用元路径挖掘用户行为间复杂的语义关系以表征新用户的属性（Yu等^[52],2019）	MF
	冷启动	提出混合图神经网络模型改善冷启动问题（Zhong等^[67],2020）	GNN
	序列推荐	提出一种基于注意力机制的循环神经网络实现序列推荐（Xia等^[63],2017）	RNN
		提出改进的基于图的潜在表示模型来捕获时间序列影响（Lu等^[58],2020）	Hybrid
		利用循环神经网络学习情景信息来预测用户下一个签到的兴趣点（Wang等^[42],2021）	RNN
	动态推荐	提出兴趣点的受欢迎程度受时间影响（Yuan等^[8],2013）	Hybrid
		考虑用户偏好的动态变化特征（Gao等^[14],2013;Ying等^[37],2019;Ma等^[38],2020）	MF、RNN等
		利用协同过滤算法与模糊聚类算法捕捉用户的时空偏好（Yin等^[39],2021）	Hybrid
	个性化推荐	提出核密度估计方法描述用户的地理行为（Zhang等^[16],2015;任星怡等^[20],2017）	MF等
	个性化推荐	融合流行度与二维的核密度估计方法捕捉活跃用户的偏好（Si等^[51],2019）	—
	异地推荐	提出概率生成模型模拟用户在异地的决策行为（任星怡等^[17],2017）	—

Table 4 Analysis of research status in POI recommendation

影响因素和问题		推荐策略	推荐算法
影响因素	地理信息	考虑用户签到兴趣点的概率与物理距离相关,提出幂律分布模型（Ye等^[7],2011）	Hybrid
		考虑用户在多个中心点范围内签到,提出多中心的高斯分布模型（Cheng等^[24],2012）	MF
		提出核密度估计方法建模用户的地理偏好（Zhang等^[15],2013）	Hybrid
		提出固定带宽核密度估计方法构建地理偏好模型（Zhang等^[16],2015）	—
		利用自适应核密度估计方法捕捉用户的地理偏好（任星怡等^[20],2017;Si等^[51],2019）	MF等
	用户偏好	将情景信息融入矩阵分解框架（任星怡等^[20],2017;彭宏伟等^[21],2019）	MF
		采用基于用户的协同过滤方法建模签到信息和地理信息（Song等^[18],2019）	CF
		将好友重要性与签到相关性融入协同过滤方法（Zhou等^[19],2019）	CF
		利用卷积神经网络挖掘用户评论中的语义信息（Xing等^[23],2019）	CNN
	社交	利用概率矩阵分解模型建模社交关系（Qian等^[30],2014）	MF
		提出基于朋友的协同过滤方法（彭宏伟等^[21],2019）	CF
		将改进后的直接信任与间接信任融入矩阵分解算法（Xu等^[33],2021）	MF
		利用图嵌入的方法表示用户间的社交关系（Zhu等^[34],2019;Chen等^[36],2021）	GE
		利用图神经网络构建社交网络图（Zhang等^[68],2021）	GNN
	时间	提出基于时间感知的兴趣点推荐模型（Yuan等^[8],2013）	Hybrid
		提出时间的差异性和连续性特征（Gao等^[14],2013）	MF
		提出基于非对称投影的时间感知嵌入方法（Ying等^[37],2019）	—
		利用时间信息捕捉用户的动态偏好（Ma等^[38],2020;Yin等^[39],2021）	RNN、Hybrid
		利用时间序列数据实现下一个兴趣点推荐（Chen等^[64],2020;Huang等^[66],2021）	RNN、Hybrid
	流行度	考虑兴趣点流行度影响用户的决策行为（Yuan等^[8],2013;任星怡等^[20],2017）	Hybrid、MF
		利用文本和图像信息预测兴趣点流行度（Yang等^[49],2019）	—
		将流行度特征结合二维的核密度估计与一维的幂律分布（Si等^[51],2019）	—
	内容	将兴趣点的文本内容嵌入深度学习模型（Chen等^[44],2020）	Hybrid
	内容	采用加权矩阵分解算法融合图像内容和地理信息（Zhang等^[46],2019）	MF
推荐问题	数据稀疏性	利用矩阵分解技术缓解稀疏问题（Gao等^[14],2013;温彦等^[25],2019）	MF
		利用图嵌入表示方法缓解稀疏性（Zhu等^[34],2019）	GE
		采用混合图卷积网络应对稀疏性问题（Zhong等^[67],2020）	GNN
	冷启动	利用元路径挖掘用户行为间复杂的语义关系以表征新用户的属性（Yu等^[52],2019）	MF
	冷启动	提出混合图神经网络模型改善冷启动问题（Zhong等^[67],2020）	GNN
	序列推荐	提出一种基于注意力机制的循环神经网络实现序列推荐（Xia等^[63],2017）	RNN
		提出改进的基于图的潜在表示模型来捕获时间序列影响（Lu等^[58],2020）	Hybrid
		利用循环神经网络学习情景信息来预测用户下一个签到的兴趣点（Wang等^[42],2021）	RNN
	动态推荐	提出兴趣点的受欢迎程度受时间影响（Yuan等^[8],2013）	Hybrid
		考虑用户偏好的动态变化特征（Gao等^[14],2013;Ying等^[37],2019;Ma等^[38],2020）	MF、RNN等
		利用协同过滤算法与模糊聚类算法捕捉用户的时空偏好（Yin等^[39],2021）	Hybrid
	个性化推荐	提出核密度估计方法描述用户的地理行为（Zhang等^[16],2015;任星怡等^[20],2017）	MF等
	个性化推荐	融合流行度与二维的核密度估计方法捕捉活跃用户的偏好（Si等^[51],2019）	—
	异地推荐	提出概率生成模型模拟用户在异地的决策行为（任星怡等^[17],2017）	—

Table 5 Summary of evaluation metrics of several algorithms

代表算法	precision	recall	F1	MAE	MAP	NDCG	时间复杂度	空间复杂度
LRT^[14]	√	√					$O (m n d)$	N/A
ASMF^[56]	√	√			√		$O (m n d 2)$	N/A
UCGSMF^[21]	√	√					$O ((m + n) d 2 + M d)$	N/A
GSBPR^[74]	√	√			√	√	$O (T m d)$	N/A
APRA-SA^[51]	√	√	√	√			$O (m d + n 2)$	N/A
SPR^[45]				√			$O (T n 3 d)$	$O (m n + n 2)$
MANC^[75]	√	√			√	√	N/A	N/A
ATST-LSTM^[41]	√	√	√				$O (m n d 2)$	N/A
Loc-Interest-LSTM^[64]	√	√	√				$O (T K n)$	$O (K n)$

Table 5 Summary of evaluation metrics of several algorithms

代表算法	precision	recall	F1	MAE	MAP	NDCG	时间复杂度	空间复杂度
LRT^[14]	√	√					$O (m n d)$	N/A
ASMF^[56]	√	√			√		$O (m n d 2)$	N/A
UCGSMF^[21]	√	√					$O ((m + n) d 2 + M d)$	N/A
GSBPR^[74]	√	√			√	√	$O (T m d)$	N/A
APRA-SA^[51]	√	√	√	√			$O (m d + n 2)$	N/A
SPR^[45]				√			$O (T n 3 d)$	$O (m n + n 2)$
MANC^[75]	√	√			√	√	N/A	N/A
ATST-LSTM^[41]	√	√	√				$O (m n d 2)$	N/A
Loc-Interest-LSTM^[64]	√	√	√				$O (T K n)$	$O (K n)$

References 79

[1]	BARAL R, ZHU X L, IYENGAR S S, et al. ReEL: review aware explanation of location recommendation[C]// Procee-dings of the 26th Conference on User Modeling, Adapta-tion and Personalization, Singapore, Jul 8-11, 2018. New York: ACM, 2018: 23-32.
[2]	RAVI L, SUBRAMANIYASWAMY V, VARADHARAJAN V, et al. Efficient user profiling based intelligent travel reco-mmender system for individual and group of users[J]. Mobile Networks and Applications, 2019, 24(3): 1018-1033. DOI URL
[3]	KOREN Y, BELL R, VOLINSKY C. Matrix factorization techniques for recommender systems[J]. Computer, 2009, 42(8): 30-37.
[4]	ZHANG Z P, ZHANG Y, REN Y G. Employing neighborhood reduction for alleviating sparsity and cold start problems in user-based collaborative filtering[J]. Information Retrieval Journal, 2020, 23(4): 449-472. DOI URL
[5]	XUE F, HE X N, WANG X, et al. Deep item-based collabo-rative filtering for top-n recommendation[J]. ACM Transac-tions on Information Systems, 2019, 37(3): 1-25.
[6]	郭宁宁, 王宝亮, 侯永宏, 等. 融合社交网络特征的协同过滤推荐算法[J]. 计算机科学与探索, 2018, 12(2): 208-217.
	GUO N N, WANG B L, HOU Y H, et al. Collaborative filtering recommendation algorithm based on characteristics of social network[J]. Journal of Frontiers of Computer Science and Technology, 2018, 12(2): 208-217.
[7]	YE M, YIN P F, LEE W C. et al. Exploiting geographical influence for collaborative point-of-interest recommendation[C]// Proceedings of the 34th International ACM SIGIR Con-ference on Research and Development in Information Re-trieval, Beijing, Jul 25-29, 2011. New York: ACM, 2011: 325-334.
[8]	YUAN Q, CONG G, MA Z Y, et al. Time-aware point-of-interest recommendation[C]// Proceedings of the 36th Inter-national ACM SIGIR Conference on Research and Deve-lopment in Information Retrieval, Dublin, Jul 28-Aug 1, 2013. New York: ACM, 2013: 363-372.
[9]	ADOMAVICIUS G, TUZHILIN A. Toward the next genera-tion of recommender systems: a survey of the state-of-the-art and possible extensions[J]. IEEE Transactions on Know-ledge and Data Engineering, 2005, 17(6): 734-749.
[10]	SHIH D H, YEN D C, LIN H C, et al. An implementation and evaluation of recommender systems for traveling abroad[J]. Expert Systems with Applications, 2011, 38(12): 15344-15355.
[11]	KARDAN A A, EBRAHIMI M. A novel approach to hybrid recommendation systems based on association rules mining for content recommendation in asynchronous discussion groups[J]. Information Sciences, 2013, 219: 93-110. DOI URL
[12]	GUO B, LI J, ZHENG V W, et al. CityTransfer: transferring inter-and intra-city knowledge for chain store site recom-mendation based on multi-source urban data[J]. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2018, 1(4): 1-23.
[13]	焦旭, 肖迎元, 郑文广, 等. 基于位置的社交网络推荐技术研究进展[J]. 计算机研究与发展, 2018, 55(10): 2291-2306.
	JIAO X, XIAO Y Y, ZHENG W G, et al. Research progress of recommendation technology in location-based social net-works[J]. Journal of Computer Research and Development, 2018, 55(10): 2291-2306.
[14]	GAO H J, TANG J L, HU X, et al. Exploring temporal effects for location recommendation on location-based social networks[C]// Proceedings of the 7th ACM Conference on Recommender Systems, Hong Kong, China, Oct 12-16, 2013. New York: ACM, 2013: 93-100.
[15]	ZHANG J D, CHOW C Y. iGSLR: personalized geo-social location recommendation—a kernel density estimation app-roach[C]// Proceedings of the 21st SIGSPATIAL International Conference on Advances in Geographic Information Systems, Orlando, Nov 5-8, 2013. New York: ACM, 2013: 324-333.
[16]	ZHANG J D, CHOW C Y. CoRe: exploiting the personalized influence of two-dimensional geographic coordinates for location recommendation[J]. Journal of Information Sciences, 2015, 293(1): 163-181. DOI URL
[17]	任星怡, 宋美娜, 宋俊德. 基于用户签到行为的兴趣点推荐[J]. 计算机学报, 2017, 40(1):28-51.
	REN X Y, SONG M N, SONG J D. Point-of-interest reco-mmendation based on the user check-in behavior[J]. Chinese Journal of Computers, 2017, 40(1):28-51.
[18]	SONG C, WEN J H, LI S. Personalized POI recommenda-tion based on check-in data and geographical regional influ-ence[C]// Proceedings of the 3rd International Conference on Machine Learning and Soft Computing, Da Lat, Jan 25-28, 2019. New York: ACM, 2019: 128-133.
[19]	ZHOU J J, LIU B, CHEN Y F, et al. UFC: a unified POI recommendation framework[J]. Arabian Journal for Science and Engineering, 2019, 44(11): 9321-9332. DOI URL
[20]	任星怡, 宋美娜, 宋俊德. 基于位置社交网络的上下文感知的兴趣点推荐[J]. 计算机学报, 2017, 40(4): 824-841.
	REN X Y, SONG M N, SONG J D. Context-aware point-of-interest recommendation in location-based social networks[J]. Chinese Journal of Computers, 2017, 40(4): 824-841.
[21]	彭宏伟, 靳远远, 吕晓强, 等. 一种基于矩阵分解的上下文感知POI推荐算法[J]. 计算机学报, 2019, 42(8): 1797-1811.
	PENG H W, JIN Y Y, LV X Q, et al. Context-aware POI recommendation based on matrix factorization[J]. Chinese Journal of Computers, 2019, 42(8): 1797-1811.
[22]	XIONG X, QIAO S J, HAN N, et al. Where to go: an effective point-of-interest recommendation framework for heterogeneous social networks[J]. Neurocomputing, 2020, 373: 56-69. DOI URL
[23]	XING S N, LIU F A, WANG Q Q, et al. Content-aware point-of-interest recommendation based on convolutional neural network[J]. Applied Intelligence, 2019, 49(3): 858-871. DOI URL
[24]	CHENG C, YANG H Q, KING I, et al. Fused matrix facto-rization with geographical and social influence in location-based social networks[C]// Proceedings of the 26th AAAI Conference on A.pngicial Intelligence, Toronto, Jul 22-26, 2012. Menlo Park: AAAI, 2012: 17-23.
[25]	温彦, 马立健, 曾庆田, 等. 基于地理信息偏好修正和社交关系偏好隐式分析的POI推荐[J]. 数据分析与知识发现, 2019, 3(8): 30-40.
	WEN Y, MA L J, ZENG Q T, et al. POI recommendation based on geographic and social relationship preferences[J]. Data Analysis and Knowledge Discovery, 2019, 3(8): 30-40.
[26]	夏英, 张金凤. 融合社交关系和局部地理因素的兴趣点推荐[J]. 计算机工程与应用, 2021, 57(15): 133-139.
	XIA Y, ZHANG J F. POI recommendation fusing social relations and local geographic factors[J]. Computer Enginee-ring and Applications, 2021, 57(15): 133-139.
[27]	LIU T C, LIAO J X, WU Z G, et al. Exploiting geographical-temporal awareness attention for next point-of-interest reco-mmendation[J]. Neurocomputing, 2020, 400: 227-237. DOI URL
[28]	LIAN D F, ZHAO C, XIE X, et al. GeoMF: joint geogra-phical modeling and matrix factorization for point-of-interest recommendation[C]// Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, Aug 24-27, 2014. New York: ACM, 2014: 831-840.
[29]	RAHMANI H A, ALIANNEJADI M, AHMADIAN S, et al. LGLMF: local geographical based logistic matrix facto-rization model for POI recommendation[C]// LNCS 12004: Proceedings of the 15th Asia Information Retrieval Societies Conference on Information Retrieval Technology, Hong Kong, China, Nov 7-9, 2019. Cham: Springer, 2020: 66-78.
[30]	QIAN X M, FENG H, ZHAO G S, et al. Personalized reco-mmendation combining user interest and social circle[J]. IEEE Transactions on Knowledge and Data Engineering, 2014, 26(7): 1763-1777. DOI URL
[31]	ZHANG Z Y, LIU Y, ZHANG Z J, et al. Fused matrix facto-rization with multi-tag, social and geographical influences for POI recommendation[J]. World Wide Web, 2019, 22(3): 1135-1150. DOI URL
[32]	ZHU J H, WANG C, GUO X, et al. Friend and POI reco-mmendation based on social trust cluster in location-based social networks[J]. EURASIP Journal on Wireless Commu-nications and Networking, 2019(1): 89.
[33]	XU C H, DING A S, ZHAO K D. A novel POI recomme-ndation method based on trust relationship and spatial-temporal factors[J]. Electronic Commerce Research and Applications, 2021, 48: 101060. DOI URL
[34]	ZHU J H, GUO X. Deep neural model for point-of-interest recommendation fused with graph embedding representation[C]// LNCS 11604: Proceedings of the 14th International Conference on Wireless Algorithms, Systems, and Applica-tions, Honolulu, Jun 24-26, 2019. Cham: Springer, 2019: 495-506.
[35]	QIAO Y, LUO X, LI C, et al. Heterogeneous graph-based joint representation learning for users and POIs in location-based social network[J]. Information Processing & Mana-gement, 2020, 57(2): 102151.
[36]	CHEN L, YING Y K, LYU D, et al. A multi-task embedding based personalized POI recommendation method[J]. CCF Transactions on Pervasive Computing and Interaction, 2021, 3(3): 253-269. DOI URL
[37]	YING H C, WU J, XU G D, et al. Time-aware metric em-bedding with asymmetric projection for successive POI reco-mmendation[J]. World Wide Web, 2019, 22(5): 2209-2224. DOI URL
[38]	MA Y X, GAN M X. Exploring multiple spatio-temporal information for point-of-interest recommendation[J]. Soft Computing, 2020, 24: 18733-18747. DOI URL
[39]	YIN M, LIU Y, ZHOU X, et al. A fuzzy clustering based collaborative filtering algorithm for time-aware POI recom-mendation[J]. Journal of Physics: Conference Series, 2021, 1746(1): 012037. DOI URL
[40]	LU Y S, HUANG J L. GLR: a graph-based latent representa-tion model for successive POI recommendation[J]. Future Generation Computer Systems, 2020, 102: 230-244. DOI URL
[41]	HUANG L W, MA Y T, WANG S B, et al. An attention-based spatiotemporal LSTM network for next POI recom-mendation[J]. IEEE Transactions on Services Computing, 2021, 14(6): 1585-1597. DOI URL
[42]	WANG H L, LI P Y, LIU Y, et al. Towards real-time demand-aware sequential POI recommendation[J]. Information Sciences, 2021, 547: 482-497. DOI URL
[43]	WU Y, LI K, ZHAO G, et al. Personalized long- and short-term preference learning for next POI recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2022, 34(4): 1944-1957. DOI URL
[44]	CHEN L, ZHANG L, CAO S S, et al. Personalized itinerary recommendation: deep and collaborative learning with textual information[J]. Expert Systems with Applications, 2020, 144: 113070. DOI URL
[45]	ZHAO G, LOU P, QIAN X, et al. Personalized location recommendation by fusing sentimental and spatial context[J]. Knowledge-Based Systems, 2020, 196: 105849. DOI URL
[46]	ZHANG Z B, ZOU C, DING R F, et al. VCG: exploiting visual contents and geographical influence for point-of-interest recommendation[J]. Neurocomputing, 2019, 357: 53-65. DOI URL
[47]	CONTRACTOR D, GOEL S, SINGLA P. Joint spatio-textual reasoning for answering tourism questions[C]// Pro-ceedings of the 2021 Web Conference, Ljubljana, Apr 19-23, 2021. New York: ACM, 2021: 1978-1989.
[48]	SÁNCHEZ VILLEGAS D, ALETRAS N. Point-of-interest type prediction using text and images[J]. arXiv: 2109. 00602, 2021.
[49]	YANG Y, DUAN Y Q, WANG X Z, et al. Hierarchical multi-clue modelling for POI popularity prediction with hetero-geneous tourist information[J]. IEEE Transactions on Know-ledge and Data Engineering, 2019, 31(4): 757-768.
[50]	YAO Z J, FU Y J, LIU B, et al. POI recommendation: a temporal matching between POI popularity and user regu-larity[C]// Proceedings of the IEEE 16th International Confer-ence on Data Mining, Barcelona, Dec 12-15, 2016. Washing-ton: IEEE Computer Society, 2016: 549-558.
[51]	SI Y L, ZHANG F Z, LIU W Y. An adaptive point-of-interest recommendation method for location-based social networks based on user activity and spatial features[J]. Knowledge-Based Systems, 2019, 163: 267-282. DOI URL
[52]	YU D J, XU K H, WANG D J, et al. Point-of-interest recommendation based on user contextual behavior semantics[J]. International Journal of Software Engineering and Know-ledge Engineering, 2019, 29(11/12): 1781-1799.
[53]	DAVTALAB M, ALESHEIKH A A. A POI recommendation approach integrating social spatio-temporal information into probabilistic matrix factorization[J]. Knowledge and Infor-mation Systems, 2021, 63(1): 65-85.
[54]	YU D J, WANYAN W B, WANG D J. Leveraging contextual influence and user preferences for point-of-interest recom-mendation[J]. Multimedia Tools and Applications, 2021, 80(1): 1487-1501. DOI URL
[55]	YUAN H, XU J, ZHENG N, et al. PRPOIR: exploiting the region-level interest for POI recommendation[C]// Procee-dings of the 32nd IEEE International Conference on Tools with A.pngicial Intelligence, Baltimore, Nov 9-11, 2020. Pis-cataway: IEEE, 2020: 59-66.
[56]	LI H, GE Y, HONG R C, et al. Point-of-interest recom-mendations: learning potential check-ins from friends[C]// Proceedings of the 22nd ACM SIGKDD International Con-ference on Knowledge Discovery and Data Mining, San Francisco, Aug 13-17, 2016. New York: ACM, 2016: 975-984.
[57]	RAHMANI H A, ALIANNEJADI M, BARATCHI M, et al. Joint geographical and temporal modeling based on matrix factorization for point-of-interest recommendation[J]. Advan-ces in Information Retrieval, 2020, 12035: 205-219.
[58]	LU Y S, HUANG J L. GLR: a graph-based latent representa-tion model for successive POI recommendation[J]. Future Generation Computer Systems, 2020, 102: 230-244. DOI URL
[59]	HU X J, XU J J, WANG W Q, et al. A graph embedding based model for fine-grained POI recommendation[J]. Neuro-computing, 2021, 428: 376-384.
[60]	CHRISTOFORIDIS G, KEFALAS P, PAPADOPOULOS A N, et al. RELINE: point-of-interest recommendations using multiple network embeddings[J]. Knowledge and Information Systems, 2021, 63(4): 791-817. DOI URL
[61]	XING S, LIU F, ZHAO X, et al. Points-of-interest recom-mendation based on convolution matrix factorization[J]. Applied intelligence, 2018, 48(8): 2458-2469. DOI URL
[62]	冯浩, 黄坤, 李晶, 等. 基于深度学习的混合兴趣点推荐算法[J]. 电子与信息学报, 2019, 41(4): 880-887.
	FENG H, HUANG K, LI J, et al. Hybrid point of interest recommendation algorithm based on deep learning[J]. Journal of Electronics & Information Technology, 2019, 41(4): 880-887.
[63]	XIA B, LI Y, LI Q M, et al. Attention-based recurrent neural network for location recommendation[C]// Proceedings of the 12th International Conference on Intelligent Systems and Knowledge Engineering, Nanjing, Nov 24-26, 2017. Piscataway: IEEE, 2017: 1-6.
[64]	CHEN M, LI W Z, QIAN L, et al. Next POI recommenda-tion based on location interest mining with recurrent neural networks[J]. Journal of Computer Science and Technology, 2020, 35(3): 603-616. DOI URL
[65]	SUN K, QIAN T, CHEN T, et al. Where to go next: modeling long- and short-term user preferences for point-of-interest recommendation[C]// Proceedings of the 34th AAAI Confer-ence on A.pngicial Intelligence, the 32nd Innovative Appli-cations of A.pngicial Intelligence Conference, the 10th AAAI Symposium on Educational Advances in A.pngicial Intellig-ence, New York, Feb 7-12, 2020. Menlo Park: AAAI, 2020: 214-221.
[66]	HUANG C M, WU C Y. The point of interest (POI) recom-mendation for mobile digital culture heritage (M-DCH) based on the behavior analysis using the recurrent neural networks (RNN) and user-collaborative filtering[J]. Journal of Internet Technology, 2021, 22(4): 821-833. DOI URL
[67]	ZHONG T, ZHANG S, ZHOU F, et al. Hybrid graph convo-lutional networks with multi-head attention for location recommendation[J]. World Wide Web, 2020, 23(6): 3125-3151. DOI URL
[68]	ZHANG J, LIU X, ZHOU X, et al. Leveraging graph neural networks for point-of-interest recommendations[J]. Neuro-computing, 2021, 462: 1-13.
[69]	KANG Y, CHEN J, CAO Y, et al. A higher-order graph convolutional network for location recommendation of an air-quality-monitoring station[J]. Remote Sensing, 2021, 13(8): 1600. DOI URL
[70]	HUANG L, MA Y, LIU Y, et al. DAN-SNR: a deep attentive network for social-aware next point-of-interest recommen-dation[J]. ACM Transactions on Internet Technology, 2020, 21(1): 1-27.
[71]	LUO Y T, LIU Q, LIU Z C. STAN: spatio-temporal attention network for next location recommendation[C]// Proceedings of the 2021 Web Conference, Ljubljana, Apr 19-23, 2021. New York: ACM, 2021: 2177-2185.
[72]	张松慧, 熊汉江. 融合神经网络和泊松分解的兴趣点推荐算法[J]. 计算机工程与应用, 2020, 56(21): 176-186.
	ZHANG S H, XIONG H J. Point-of-interest recommenda-tion algorithm based on Poisson factorization and neural network[J]. Computer Engineering and Applications, 2020, 56(21): 176-186.
[73]	XIE M, YIN H Z, WANG H, et al. Learning graph-based POI embedding for location-based recommendation[C]// Proceedings of the 25th ACM International Conference on Information and Knowledge Management, Indianapolis, Oct 24-28, 2016. New York: ACM, 2016: 15-24.
[74]	GAO R, LI J, DU B, et al. Exploiting geo-social correlations to improve pairwise ranking for point-of-interest recom-mendation[J]. China Communications, 2018, 15(7): 180-201. DOI URL
[75]	CHANG L, CHEN W, HUANG J B, et al. Exploiting multi-attention network with contextual influence for point-of-interest recommendation[J]. Applied Intelligence, 2021, 51(4): 1904-1917. DOI URL
[76]	ZHU J H, GUO X. Deep neural model for point-of-interest recommendation fused with graph embedding representa-tion[C]// LNCS 11604: Proceedings of the 14th International Conference on Wireless Algorithms, Systems, and Applica-tions, Honolulu, Jun 24-26, 2019. Cham: Springer, 2019: 495-506.
[77]	GAO H J, TANG J L, HU X, et al. Content-aware point of interest recommendation on location-based social networks[C]// Proceedings of the 29th AAAI Conference on A.pngicial Intelligence, Austin, Jan 25-30, 2015. Menlo Park: AAAI, 2015: 1721-1727.
[78]	CHO E, MYERS S A, LESKOVEC J. Friendship and mobility: user movement in location-based social networks[C]// Pro-ceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Diego, Aug 21-24, 2011. New York: ACM, 2011: 1082-1090.
[79]	YELP. Challenge data set[EB/OL]. (2014-04-25)[2021-10-25]. .