TPR⁃TF:基于张量分解的时间敏感兴趣点推荐模型

doi:10.13229/j.cnki.jdxbgxb20180193

摘要/Abstract

摘要：

针对现有的兴趣点(POI)推荐研究没有合理地利用POI推荐的时间敏感性，对用户在不同时间段的行为偏好没有给予充分考虑，造成推荐效果较差的问题,提出了一种基于层次聚类的时间动态分段算法。把时间敏感的推荐和用户的直接朋友及潜在朋友影响相结合，扩充了用户的社交影响范围。在模型学习过程中采用按访问频次分布随机选择POI位置的方法，改善了经典的贝叶斯个性化排序（BPR）方法。实验结果表明，本文模型性能优于目前的主流POI推荐模型。

关键词: 计算机应用, 兴趣点推荐, 张量分解, 时间敏感, 社交关系

Abstract:

With the rapid growth of the location?based social networks, Point of Interest (POI) recommen?dation has become an important research topic in the field of data mining. Existing approaches for POI recommendation task do not reasonably utilize the time sensitivity of POI recommendations and have not taken full account of the user's behavior preferences at different time periods, causing the POI recommendation performance is poor. Firstly, this paper studies the POI recommendation problem of time sensitivity and proposes a time dynamic partition algorithm based on hierarchical clustering. Through partition the fine grain of time, the result of the experiment is more reasonable and effective than the previous experiments which partition time is evenly given by experience. Secondly, by combining the time?aware recommendation with the influence of the user's direct friendship and potential friendship, the paper expands the scope of user's social influence, and then further improves the POI recommendation performance. Lastly, using the method of randomly selecting POIs by the frequency distribution of check?ins, it improves the classic BPR method. Experimental results on the two datasets indicate that the TPR?TF model is superior to the current mainstream POI recommendation models, in terms of precision and recall.

Key words: computer application, point of interest(POI) recommendation, tensor factorization, time?aware, social relationship

中图分类号:

TP391

王楠,李金宝,刘勇,张玉杰,钟颖莉. TPR⁃TF:基于张量分解的时间敏感兴趣点推荐模型[J]. 吉林大学学报(工学版), 2019, 49(3): 920-933.

Nan WANG,Jin⁃bao LI,Yong LIU,Yu⁃jie ZHANG,Ying⁃li ZHONG. TPR⁃TF： time⁃aware point of interest recommendation model based on tensor factorization[J]. Journal of Jilin University(Engineering and Technology Edition), 2019, 49(3): 920-933.

图/表 9

表1

符号表"

符号	描述
U、T、L	分别表示用户集合、时间戳集合、POI位置集合。
D	$D = {d i = (u, t, l) (u, t, l) ∈ U × T × L}$ ， $D$ 为全体签到记录集合，其中 $d i$ 为用户?时间?位置三维张量上的第 $i$ 条签到记录， $(u, t, l)$ 为用户 $u$ 在时间 $t$ 在位置 $l$ 的一条签到记录。
U _u、 L _l、 T _t	U _u、 L _l、 T _t ∈ R ^k 分别为用户 $u$ 的特征向量、POI位置 $l$ 的特征向量、时间戳 $t$ 所在时间分段的特征向量， $k$ 为特征向量的维度。
$f (u, t, l)$	用户 $u$ 在时间 $t$ 访问位置 $l$ 的可能的打分函数。
$w i, j s, w i, j c, w i, j$	$w i, j s$ 为用户 $i$ 和其朋友基于共同朋友的相似性； $w i, j c$ 为用户 $i$ 和其朋友 $j$ 基于共同签到的相似性； $w i, j$ 为用户 $i$ 和其朋友 $j$ 基于共同朋友和共同签到的相似性。
$A (i)$	表示用户 $i$ 访问过的所有POI位置的集合。
$l i > u, t l j$	对于 $? l i, l j ∈ L$ ,用户 $u$ 在时间t优先选择 $l i$ 而非 $l j$ 的一种顺序关系。
L _soc?fri、L _{soc?undirect?fri}	分别为直接朋友关系的社交影响正则化项、潜在朋友关系的社交影响正则化项。
θ、γ、μ、λ	分别为参数集合、步长、朋友正则化项系数以及与 U _u 、 L _l 、 T _t 有关的正则化项系数。

表1

表2

图1

表3

表4

图2

图3

图4

图5

参考文献 25

1	Ye M , Yin P F , Lee W , et al . Exploiting geographical influence for collaborative point⁃of⁃interest recommendation[C]∥Proceedings of the 34th International ACM SIGIR Conference on Research and Development in Information Retrieval, Beijing, China, 2011: 325⁃334.
2	Yuan Q , Cong G , Ma Z Y , et al . Time⁃aware point⁃of⁃interest recommendation[C]∥Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval, Dublin, Ireland, 2013: 363⁃372.
3	Bao J , Zheng Y , Mokbel M , et al . Location⁃based and preference⁃aware recommendation using sparse geo⁃social networking data[C]∥Proceedings of the 20th International Conference on Advances in Geographic Information Systems, Redondo Beach, USA, 2012: 199⁃208.
4	Lian D , Zhao C , Xie X , et al . GeoMF: joint geographical 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, USA, 2014: 831⁃840.
5	Liu S D , Meng X W . Recommender systems in location⁃based social networks[J]. Chinese Journal of Computers, 2015, 38(2): 322⁃336.
6	Goodfellow I , Bengio Y , Courville A , et al . Deep Learning[M]. Cambridge: MIT press, 2016.
7	Yao Z , Fu Y , Liu B , et al . POI recommendation: a temporal matching between POI popularity and user regularity[C]∥Proceedings of International Conference on Data Mining, Barcelona, Spain, 2017: 549⁃558.
8	Cho E , Myers S A , Leskovec J . Friendship and mobility: user movement in location⁃based social networks[C]∥Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Diego, USA, 2011: 1082⁃1090.
9	Gao H , Tang J , 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, 2013: 93⁃100.
10	Yuan Q , Cong G , Sun A . Graph⁃based point⁃of⁃interest recommendation with geographical and temporal influences[C]∥Proceedings of the 23rd ACM International Conference on Conference on Information and Knowledge Management, Shanghai, China, 2014: 659⁃668.
11	Hosseini S , Li L T . Point⁃of⁃interest recommendation using temporal orientations of users and locations[C]∥Proceedings of the 21st International Conference on Database Systems for Advanced Applications, Dallas, USA, 2016: 330⁃347.
12	Ma H , Zhou D , Liu C , et al . Recommender systems with social regularization[C]∥Proceedings of the Forth International Conference on Web Search and Web Data Mining, Kowloon, Hong Kong, 2011: 287⁃296.
13	Zhang J , Chow C . GeoSoCa: exploiting geographical, social and categorical correlations for point⁃of⁃interest recommendations[C]∥Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval, Santiago, Chile, 2015: 443⁃452.
14	Jamali M , Ester M . A matrix factorization technique with trust propagation for recommendation in social networks[C]∥Proceedings of the Fourth ACM Conference on Recommender Systems, Barcelona, Spain, 2010:135⁃142.
15	Ma H , Zhou T C , Lyu M R , et al . Improving recommender systems by incorporating social contextual information[J]. ACM Transactions on Information Systems, 2011, 29(2):1⁃23.
16	Cheng C , Yang H , King I , et al . Fused matrix factorization with geographical and social influence in location⁃based social networks[C]∥Proceedings of the National Conference on Artificial Intelligence, Toronto, Canada, 2012: 17⁃23.
17	Rendle S , Freudenthaler C , Gantner Z , et al . BPR: Bayesian personalized ranking from implicit feedback[C]∥Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, Montreal, Canada, 2009: 452⁃461.
18	Zhao S L , King I , Lyu M R . A survey of point⁃of⁃interest recommendation in location⁃based social networks[J/OL].[2018⁃02⁃26].https:∥⁃of⁃interest_Recommendation_in_Location⁃based_Social_Networks/links/5787365a08ae36ad40a6a4e8 /A⁃Survey⁃of⁃Point⁃of⁃interest⁃Recommendation⁃in⁃Location⁃based⁃Social⁃Networks.pdf?origin=publication_detail.
19	Ye M , Yin P , Lee W C . Location recommendation for location⁃based social networks[C]∥Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems, San Jose, USA, 2010: 458⁃461.
20	Zhang W , Wang J . Location and time aware social collaborative retrieval for new successive point⁃of⁃interest recommendation[C]∥Proceedings of the 24th ACM International Conference on Information and Knowledge Management, Melbourne, Australia, 2015(19⁃23): 1221⁃1230.
21	Ye M , Liu X , Lee W C . Exploring social influence for recommendation: a generative model approach[C]∥Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval, Portland, USA, 2012: 671⁃680.
22	Li H , Ge Y , Zhu H , et al . Point⁃of⁃interest recommendations: learning potential check⁃ins from friends[C]∥Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, USA, 2016(13⁃17): 975⁃984.
23	Zhang J , Chow C , Li Y , et al . LORE: exploiting sequential influence for location recommendations[C]∥Proceedings of the 22nd SIGSPATIAL International Conference on Advances in Geographic Information Systems, Dallas, Fort Worth, USA, 2014(4⁃7): 103⁃112.
24	Yao L , Sheng Q Z , Qin Y , et al . Context⁃aware point⁃of⁃interest recommendation using tensor factorization with social regularization[C]∥Proceedings of the 38th International ACM SIGIR conference on Research and Development in Information Retrieval, Santiago, Chile, 2015: 1007⁃1010.
25	Li X , Cong G , Li X L , et al . Rank⁃GeoFM: a ranking based geographical factorization method for point of interest recommendation[C]∥Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval, Santiago, Chile, 2015: 433⁃442.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

类别	Gowalla	Brightkite
用户数	26 287	24 227
POI数	11 640	43 592
签到记录数	1 104 464	3 260 809

序号	常规时间分段	时间动态分段
T1	00:00~06:00	00:08:55~06:57:60
T2	06:00~12:00	06:57:60~13:24:03
T3	12:00~18:00	13:24:03~19:00:01
T4	18:00~00:00	19:00:01~00:08:55

序号	常规时间分段	时间动态分段
T1?工作日	00:00~06:00	00:04:40~05:36:15
T2?工作日	06:00~12:00	05:36:15~10:41:59
T3?工作日	12:00~18:00	10:41:59~16:16:48
T4?工作日	18:00~00:00	16:16:48~00:04:40
T5?非工作日	00:00~06:00	04:02:08~07:48:44
T6?非工作日	06:00~12:00	07:48:44~12:43:10
T7?非工作日	12:00~18:00	12:43:10~16:13:57
T8?非工作日	18:00~00:00	16:13:57~04:02:08

[1]	刘富,宗宇轩,康冰,张益萌,林彩霞,赵宏伟. 基于优化纹理特征的手背静脉识别系统[J]. 吉林大学学报(工学版), 2018, 48(6): 1844-1850.
[2]	王利民,刘洋,孙铭会,李美慧. 基于Markov blanket的无约束型K阶贝叶斯集成分类模型[J]. 吉林大学学报(工学版), 2018, 48(6): 1851-1858.
[3]	金顺福,王宝帅,郝闪闪,贾晓光,霍占强. 基于备用虚拟机同步休眠的云数据中心节能策略及性能[J]. 吉林大学学报(工学版), 2018, 48(6): 1859-1866.
[4]	赵东,孙明玉,朱金龙,于繁华,刘光洁,陈慧灵. 结合粒子群和单纯形的改进飞蛾优化算法[J]. 吉林大学学报(工学版), 2018, 48(6): 1867-1872.
[5]	刘恩泽,吴文福. 基于机器视觉的农作物表面多特征决策融合病变判断算法[J]. 吉林大学学报(工学版), 2018, 48(6): 1873-1878.
[6]	欧阳丹彤, 范琪. 子句级别语境感知的开放信息抽取方法[J]. 吉林大学学报(工学版), 2018, 48(5): 1563-1570.
[7]	刘富, 兰旭腾, 侯涛, 康冰, 刘云, 林彩霞. 基于优化k-mer频率的宏基因组聚类方法[J]. 吉林大学学报(工学版), 2018, 48(5): 1593-1599.
[8]	桂春, 黄旺星. 基于改进的标签传播算法的网络聚类方法[J]. 吉林大学学报(工学版), 2018, 48(5): 1600-1605.
[9]	刘元宁, 刘帅, 朱晓冬, 陈一浩, 郑少阁, 沈椿壮. 基于高斯拉普拉斯算子与自适应优化伽柏滤波的虹膜识别[J]. 吉林大学学报(工学版), 2018, 48(5): 1606-1613.
[10]	车翔玖, 王利, 郭晓新. 基于多尺度特征融合的边界检测算法[J]. 吉林大学学报(工学版), 2018, 48(5): 1621-1628.
[11]	赵宏伟, 刘宇琦, 董立岩, 王玉, 刘陪. 智能交通混合动态路径优化算法[J]. 吉林大学学报(工学版), 2018, 48(4): 1214-1223.
[12]	黄辉, 冯西安, 魏燕, 许驰, 陈慧灵. 基于增强核极限学习机的专业选择智能系统[J]. 吉林大学学报(工学版), 2018, 48(4): 1224-1230.
[13]	傅文博, 张杰, 陈永乐. 物联网环境下抵抗路由欺骗攻击的网络拓扑发现算法[J]. 吉林大学学报(工学版), 2018, 48(4): 1231-1236.
[14]	曹洁, 苏哲, 李晓旭. 基于Corr-LDA模型的图像标注方法[J]. 吉林大学学报(工学版), 2018, 48(4): 1237-1243.
[15]	侯永宏, 王利伟, 邢家明. 基于HTTP的动态自适应流媒体传输算法[J]. 吉林大学学报(工学版), 2018, 48(4): 1244-1253.