The evolution of the World Wide Web (WWW) and the smart-phone technologies have played a key role in the revolution of our daily life. The location-based social networks (LBSN) have emerged and facilitated the users to share the check-in information and multimedia contents. The Point of Interest (POI) recommendation system uses the check-in information to predict the most potential check-in locations. The different aspects of the check-in information, for instance, the geographical distance, the category, and the temporal popularity of a POI; and the temporal check-in trends, and the social (friendship) information of a user plays a crucial role in an efficient recommendation. In this paper, we propose a fused recommendation model termed MAPS (Multi Aspect Personalized POI Recom- mender System) which will be the rst in our knowledge to fuse the categorical, the temporal, the social and the spatial aspects in a single model. The major contribution of this paper are: (i) it realizes the problem as a graph of location nodes with constraints on the category and the distance as- pects (i.e. the edge between two locations is constrained by a threshold distance and the category of the locations), (ii) it proposes a multi-aspect fused POI recommendation model, and (iii) it extensively evaluates the model with two real-world data sets. The paper was published in ACM RecSys 2016. The paper can be accessed from the url http://dl.acm.org/citation.cfm?id=2959187.

1. MAPS: A Multi Aspect Personalized
POI Recommender System
Ramesh Baral, Tao Li
Florida International University
Miami, FL
September 18, 2016

2. Location Based Social Network
(LBSN)
2
[1] Bao, J., Zheng, Y., & Mokbel, M. F. (2012, November). Location-based and preference-aware recommendation using
sparse geo-social networking data. InProceedings of the 20th International Conference on Advances in Geographic
Information Systems (pp. 199-208). ACM.
[2] http://research.microsoft.com/en-us/projects/lbsn/

3. Introduction
• Objective:
– Efficient location recommendation
• Challenges
– LBSN rating matrix => {user-check-ins} often sparse
– Difficult to have explicit user profile (age, preferences etc.)
– 96% of people share < 10% of commonly visited places and
87% of people share nothing at all [1]
– Cold start problem
• Solution:
– Fusion of major aspects (check-in time, location category,
social relation, location distance etc.) associated with LBSN
3
[1] M. Ye, P. Yin, and W.-C. Lee, “Location recommendation for location-based social networks,” in Proceedings of the 18th
SIGSPATIAL International Conference on Advances in Geographic Information Systems. ACM, 2010, pp. 458–461.

4. Motivation
• Geographical
– Preference to near location
• Temporal
– Temporal popularity of POI
– Temporal check-in trend of user
• Categorical
– Locations with same category can be an option
• Social
– Influence of Friends/Followers-Followee
4

5. Example
5

6. Methodology
• Location as a node of a graph and the bag of
{user, check-in time} tuple as its attributes
• Personalized Page Rank[6]
– categorical and spatial constraints
• Time constraint => check-ins within a time
interval
• Social constraint=> check-ins made by friends
• Location rank computed with constraints
6

7. Example
7

8. Categorical Rank
8

9. Spatial Rank
9

10. Recommendation
10
Unified Rank:
Recommendation:

11. Dataset[1] statistics
Attributes Gowalla Weeplaces
Check-ins 36,001,959 7,658,368
Users 319,063 15,799
Locations 2,844,076 971,309
Social links (un-directional) 337,545 59,970
Location Categories 629 96
11
[1] X. Liu, Y. Liu, K. Aberer, and C. Miao, “Personalized point-of-interest recommendation by mining users’ preference
transition,” in Proceedings of the 22nd ACM international conference on Conference on information & knowledge
management. ACM, 2013, pp. 733–738.

12. Check-in trend with distance
13

13. Evaluation parameters
• 5-fold cross-validation
• Cases: (i) top 5, (ii) top 10, (iii) top 15 items
with the highest recommendation score
• α =0.85
• Categorical model =0.75 and =0.25
• Spatial model = 0.75 and =0.25
• For unified model, categorical aspect = 0.25
14

14. Results (Gowalla Dataset)
Models Precision Recall F-Score
Ye et al. [1] 0.03000 0.00120 0.00230
LBSNRank [2] 0.40900 0.00300 0.00600
Wang et al. [3] 0.10600 0.00200 0.00392
CLM 0.00633 0.00154 0.00247
SLM 0.25350 0.00973 0.01874
MAPS 0.35400 0.03100 0.05700*
16

15. Conclusion and Future direction
• Fused the (a) the geographical/spatial, (b) the
categorical, (c) the temporal and (d) the social
aspects into a POI recommendation model
• Future direction:
– Fusion of other aspects
– Other domains
18

16. Acknowledgements
• Anonymous reviewers
• RecSys committee
• National Science Foundation (NSF) grants CNS-
1126619, IIS-1213026, and CNS-1461926
• U.S. Department of Homeland Securitys VACCINE
Center under Award Number 2009-ST-061-
CI0001
• A gift award from Huawei Technologies Co.Ltd.
19

17. References
1. M. Ye, P. Yin, and W.-C. Lee, “Location recommendation for location-based social
networks,” in Proceedings of the 18th SIGSPATIAL International Conference on
Advances in Geographic Information Systems. ACM, 2010, pp. 458–461.
2. Z. Jin, D. Shi, Q. Wu, H. Yan, and H. Fan, “Lbsnrank: personalized pagerank on
location-based social networks,” in Proceedings of the 2012 ACM Conference on
Ubiquitous Computing. ACM, 2012, pp. 980–987.
3. H. Wang, M. Terrovitis, and N. Mamoulis, “Location recommendation in location-
based social networks using user check-in data,” in Proceedings of the 21st ACM
SIGSPATIAL International Conference on Advances in Geographic Information
Systems. ACM, 2013, pp. 374–383.
4. X. Liu, Y. Liu, K. Aberer, and C. Miao, “Personalized point-of-interest
recommendation by mining users’ preference transition,” in Proceedings of the
22nd ACM international conference on Conference on information & knowledge
management. ACM, 2013, pp. 733–738.
5. Q. Yuan, G. Cong, Z. Ma, A. Sun, and N. M. Thalmann, “Time-aware point-of-
interest recommendation,” in Proceedings of the 36th international ACM SIGIR
conference on Research and development in information retrieval. ACM, 2013, pp.
363–372.
6. T. H. Haveliwala. Topic-sensitive pagerank. In Proceedings of the 11th international
conference on World Wide Web, pages 517-526. ACM, 2002.
20

18. 21