Predicting Visitors Using Location-Based Social Networks

Muhammad Aamir Saleem, Felipe Soares Da Costa, Peter Dolog, Panagiotis Karras, Torben Bach Pedersen, Toon Calders

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

3 Citationer (Scopus)

Resumé

Location-based social networks (LBSN) are social networks complemented with users’ location data, such as geotagged activity data. Predicting such activities finds application in marketing, recommendation systems, and logistics management. In this paper, we exploit LBSN data to predict future visitors at given locations. We fetch the travel history of visitors by their check-ins in LBSNs and identify five features that significantly drive the mobility of a visitor towards a location: (i) historic visits, (ii) location category, (iii) time, (iv) distance, and (v) friends’ activities. We provide a visitor prediction model, CMViP, based on collective matrix factorization and influence propagation. CMViP first utilizes collective matrix factorization to map the first four features to a common latent space to find visitors having a significant potential to visit a given location. Then, it utilizes an influence-mining approach to further incorporate friends of those visitors, who are influenced by the visitors’ activities and likely to follow them. Our experiments on two real-world datasets show that our methods outperform the state of art in terms of precision and accuracy.
OriginalsprogEngelsk
Titel19th IEEE International Conference on Mobile Data Management (MDM)
Antal sider6
Vol/bind2018-June
ForlagIEEE Computer Society Press
Publikationsdato28 jun. 2018
Sider245-250
ISBN (Trykt)978-1-5386-4134-7
ISBN (Elektronisk)978-1-5386-4133-0
DOI
StatusUdgivet - 28 jun. 2018
BegivenhedIEEE International Conference on Mobile Data Management - Comwell Hvide Hus Aalborg, Aalborg, Danmark
Varighed: 26 jun. 201828 jun. 2018
Konferencens nummer: 19
http://mdmconferences.org/mdm2018/index.html

Konference

KonferenceIEEE International Conference on Mobile Data Management
Nummer19
LokationComwell Hvide Hus Aalborg
LandDanmark
ByAalborg
Periode26/06/201828/06/2018
Internetadresse
NavnIEEE International Conference on Mobile Data Management (MDM)
ISSN2375-0324

Fingerprint

Factorization
Recommender systems
Logistics
Marketing
Experiments

Citer dette

Saleem, M. A., Da Costa, F. S., Dolog, P., Karras, P., Pedersen, T. B., & Calders, T. (2018). Predicting Visitors Using Location-Based Social Networks. I 19th IEEE International Conference on Mobile Data Management (MDM) (Bind 2018-June, s. 245-250). IEEE Computer Society Press. IEEE International Conference on Mobile Data Management (MDM) https://doi.org/10.1109/MDM.2018.00043
Saleem, Muhammad Aamir ; Da Costa, Felipe Soares ; Dolog, Peter ; Karras, Panagiotis ; Pedersen, Torben Bach ; Calders, Toon. / Predicting Visitors Using Location-Based Social Networks. 19th IEEE International Conference on Mobile Data Management (MDM). Bind 2018-June IEEE Computer Society Press, 2018. s. 245-250 (IEEE International Conference on Mobile Data Management (MDM)).
@inproceedings{fcb0bf38906446ad900be0d44e3e5622,
title = "Predicting Visitors Using Location-Based Social Networks",
abstract = "Location-based social networks (LBSN) are social networks complemented with users’ location data, such as geotagged activity data. Predicting such activities finds application in marketing, recommendation systems, and logistics management. In this paper, we exploit LBSN data to predict future visitors at given locations. We fetch the travel history of visitors by their check-ins in LBSNs and identify five features that significantly drive the mobility of a visitor towards a location: (i) historic visits, (ii) location category, (iii) time, (iv) distance, and (v) friends’ activities. We provide a visitor prediction model, CMViP, based on collective matrix factorization and influence propagation. CMViP first utilizes collective matrix factorization to map the first four features to a common latent space to find visitors having a significant potential to visit a given location. Then, it utilizes an influence-mining approach to further incorporate friends of those visitors, who are influenced by the visitors’ activities and likely to follow them. Our experiments on two real-world datasets show that our methods outperform the state of art in terms of precision and accuracy.",
keywords = "Collective matrix factorization, Influence propagation, Location based Social Networks, Visitor prediction",
author = "Saleem, {Muhammad Aamir} and {Da Costa}, {Felipe Soares} and Peter Dolog and Panagiotis Karras and Pedersen, {Torben Bach} and Toon Calders",
year = "2018",
month = "6",
day = "28",
doi = "10.1109/MDM.2018.00043",
language = "English",
isbn = "978-1-5386-4134-7",
volume = "2018-June",
series = "IEEE International Conference on Mobile Data Management (MDM)",
publisher = "IEEE Computer Society Press",
pages = "245--250",
booktitle = "19th IEEE International Conference on Mobile Data Management (MDM)",
address = "United States",

}

Saleem, MA, Da Costa, FS, Dolog, P, Karras, P, Pedersen, TB & Calders, T 2018, Predicting Visitors Using Location-Based Social Networks. i 19th IEEE International Conference on Mobile Data Management (MDM). bind 2018-June, IEEE Computer Society Press, IEEE International Conference on Mobile Data Management (MDM), s. 245-250, Aalborg, Danmark, 26/06/2018. https://doi.org/10.1109/MDM.2018.00043

Predicting Visitors Using Location-Based Social Networks. / Saleem, Muhammad Aamir; Da Costa, Felipe Soares; Dolog, Peter; Karras, Panagiotis; Pedersen, Torben Bach; Calders, Toon.

19th IEEE International Conference on Mobile Data Management (MDM). Bind 2018-June IEEE Computer Society Press, 2018. s. 245-250 (IEEE International Conference on Mobile Data Management (MDM)).

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Predicting Visitors Using Location-Based Social Networks

AU - Saleem, Muhammad Aamir

AU - Da Costa, Felipe Soares

AU - Dolog, Peter

AU - Karras, Panagiotis

AU - Pedersen, Torben Bach

AU - Calders, Toon

PY - 2018/6/28

Y1 - 2018/6/28

N2 - Location-based social networks (LBSN) are social networks complemented with users’ location data, such as geotagged activity data. Predicting such activities finds application in marketing, recommendation systems, and logistics management. In this paper, we exploit LBSN data to predict future visitors at given locations. We fetch the travel history of visitors by their check-ins in LBSNs and identify five features that significantly drive the mobility of a visitor towards a location: (i) historic visits, (ii) location category, (iii) time, (iv) distance, and (v) friends’ activities. We provide a visitor prediction model, CMViP, based on collective matrix factorization and influence propagation. CMViP first utilizes collective matrix factorization to map the first four features to a common latent space to find visitors having a significant potential to visit a given location. Then, it utilizes an influence-mining approach to further incorporate friends of those visitors, who are influenced by the visitors’ activities and likely to follow them. Our experiments on two real-world datasets show that our methods outperform the state of art in terms of precision and accuracy.

AB - Location-based social networks (LBSN) are social networks complemented with users’ location data, such as geotagged activity data. Predicting such activities finds application in marketing, recommendation systems, and logistics management. In this paper, we exploit LBSN data to predict future visitors at given locations. We fetch the travel history of visitors by their check-ins in LBSNs and identify five features that significantly drive the mobility of a visitor towards a location: (i) historic visits, (ii) location category, (iii) time, (iv) distance, and (v) friends’ activities. We provide a visitor prediction model, CMViP, based on collective matrix factorization and influence propagation. CMViP first utilizes collective matrix factorization to map the first four features to a common latent space to find visitors having a significant potential to visit a given location. Then, it utilizes an influence-mining approach to further incorporate friends of those visitors, who are influenced by the visitors’ activities and likely to follow them. Our experiments on two real-world datasets show that our methods outperform the state of art in terms of precision and accuracy.

KW - Collective matrix factorization

KW - Influence propagation

KW - Location based Social Networks

KW - Visitor prediction

UR - http://www.scopus.com/inward/record.url?scp=85050822001&partnerID=8YFLogxK

U2 - 10.1109/MDM.2018.00043

DO - 10.1109/MDM.2018.00043

M3 - Article in proceeding

SN - 978-1-5386-4134-7

VL - 2018-June

T3 - IEEE International Conference on Mobile Data Management (MDM)

SP - 245

EP - 250

BT - 19th IEEE International Conference on Mobile Data Management (MDM)

PB - IEEE Computer Society Press

ER -

Saleem MA, Da Costa FS, Dolog P, Karras P, Pedersen TB, Calders T. Predicting Visitors Using Location-Based Social Networks. I 19th IEEE International Conference on Mobile Data Management (MDM). Bind 2018-June. IEEE Computer Society Press. 2018. s. 245-250. (IEEE International Conference on Mobile Data Management (MDM)). https://doi.org/10.1109/MDM.2018.00043