TY - GEN
T1 - A fast algorithm to build new users similarity list in neighbourhood-based collaborative filtering
AU - Lu, Zhigang
AU - Shen, Hong
N1 - Publisher Copyright:
© Springer Science+Business Media Singapore 2016.
PY - 2016
Y1 - 2016
N2 - Neighbourhood-based Collaborative Filtering (CF) has been applied in the industry for several decades because of its easy implementation and high recommendation accuracy. As the core of neighbourhood-based CF, the task of dynamically maintaining users’ similarity list is challenged by cold-start problem and scalability problem. Recently, several methods are presented on addressing the two problems. However, these methods require mn steps to compute the similarity list against the kNN attack, where m and n are the number of items and users in the system respectively. Observing that the k new users from the kNN attack, with enough recommendation data, have the same rating list, we present a faster algorithm, TwinSearch, to avoid computing and sorting the similarity list for each new user repeatedly to save the time. The computational cost of our algorithm is 1/125 of the existing methods. Both theoretical and experimental results show that the TwinSearch Algorithm achieves better running time than the traditional method.
AB - Neighbourhood-based Collaborative Filtering (CF) has been applied in the industry for several decades because of its easy implementation and high recommendation accuracy. As the core of neighbourhood-based CF, the task of dynamically maintaining users’ similarity list is challenged by cold-start problem and scalability problem. Recently, several methods are presented on addressing the two problems. However, these methods require mn steps to compute the similarity list against the kNN attack, where m and n are the number of items and users in the system respectively. Observing that the k new users from the kNN attack, with enough recommendation data, have the same rating list, we present a faster algorithm, TwinSearch, to avoid computing and sorting the similarity list for each new user repeatedly to save the time. The computational cost of our algorithm is 1/125 of the existing methods. Both theoretical and experimental results show that the TwinSearch Algorithm achieves better running time than the traditional method.
KW - Database applications
KW - Neighbourhood-based collaborative filtering
KW - Recommender systems
KW - Similarity computation
UR - http://www.scopus.com/inward/record.url?scp=84958049649&partnerID=8YFLogxK
U2 - 10.1007/978-981-10-0068-3_30
DO - 10.1007/978-981-10-0068-3_30
M3 - Conference contribution
AN - SCOPUS:84958049649
SN - 9789811000676
T3 - Lecture Notes in Electrical Engineering
SP - 229
EP - 236
BT - Advances in Parallel and Distributed Computing and Ubiquitous Services, UCAWSN and PDCAT 2015
A2 - Shen, Hong
A2 - Jeong, Young-Sik
A2 - Yi, Gangman
A2 - Park, James J.
PB - Springer Verlag
T2 - 4th International Conference on Ubiquitous Computing Application and Wireless Sensor Network, UCAWSN 2015
Y2 - 8 July 2015 through 10 July 2015
ER -