TY - GEN
T1 - Multiple days trip recommendation based on check-in data
AU - Liao, Heng Ching
AU - Chen, Yi Chung
AU - Lee, Chiang
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2015.
PY - 2015
Y1 - 2015
N2 - A travel recommender system can generate suggested itineraries for users based on their preferences. However, current systems are not capable of simultaneously considering trip length, distance, user requirements and preferences when making recommendations, being only equipped to consider one or two of these variables at one time. Also, to generate recommendations the system must process all attractions in the database, requiring more data access and longer processing time. We analyzed the check-in records of users and utilized a new concept of time intervals combined with a multiple days trip algorithm to produce itineraries compatible with the interests and needs of users. By applying R-tree to the travel recommender system, we reduced data access times and computation time. Lastly, we propose a trip evaluator equation that can be used to compare the strengths and weaknesses of each algorithm. Experimental results verified the effectiveness of our method.
AB - A travel recommender system can generate suggested itineraries for users based on their preferences. However, current systems are not capable of simultaneously considering trip length, distance, user requirements and preferences when making recommendations, being only equipped to consider one or two of these variables at one time. Also, to generate recommendations the system must process all attractions in the database, requiring more data access and longer processing time. We analyzed the check-in records of users and utilized a new concept of time intervals combined with a multiple days trip algorithm to produce itineraries compatible with the interests and needs of users. By applying R-tree to the travel recommender system, we reduced data access times and computation time. Lastly, we propose a trip evaluator equation that can be used to compare the strengths and weaknesses of each algorithm. Experimental results verified the effectiveness of our method.
UR - http://www.scopus.com/inward/record.url?scp=84946093296&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84946093296&partnerID=8YFLogxK
U2 - 10.1007/978-3-662-48319-0_25
DO - 10.1007/978-3-662-48319-0_25
M3 - Conference contribution
AN - SCOPUS:84946093296
SN - 9783662483183
T3 - Communications in Computer and Information Science
SP - 316
EP - 330
BT - Multidisciplinary Social Networks Research - 2nd International Conference, MISNC 2015, Proceedings
A2 - Wang, Kai
A2 - Uesugi, Shiro
A2 - Wang, Leon
A2 - Okuhara, Koji
A2 - Ting, I-Hsien
PB - Springer Verlag
T2 - 2nd International Conference on Multidisciplinary Social Networks Research, MISNC 2015
Y2 - 1 September 2015 through 3 September 2015
ER -