Title: A novel e-commerce customer continuous purchase recommendation model research based on colony clustering
Authors: Qibei Lu; Feipeng Guo
School of Science and Technology, Zhejiang International Study University, Hangzhou, Zhejiang Province, China
Department of Information Technology, Zhejiang Economic and Trade Polytechnic, Hangzhou, Zhejiang Province, China
Abstract: Customer purchasing behaviour in e-commerce platform has become uncertainty and jump affected by the contexts. Existing personalised recommendation models failed to deal with the problem well and they cause loss of customers constantly. This paper puts forward a novel model based on ant colony clustering algorithm to improve customers' continuous purchase intention, including for customer interest is not drift and interest has already shifting. Firstly, for the research field of e-commerce, it gives definition and structured expression to contexts connotation. Secondly, for the problem that growing data sparseness in recommendation system, it introduces ant colony algorithm to cluster similar users in order to reduce the number of candidate neighbour sets and user similarity computing time. Thus, it can improve the target users of nearest neighbour search accuracy. On this basis, according to customer interest variation characteristics, we put forward the dynamic collaborative filtering recommendation algorithm based on Maslow to do adaptive recommend. Finally, the experiment shows effectiveness of this model and the method can solve the problem of e-commerce platform customers' continuous purchase problem.
Keywords: online shopping; continuous purchase intention; contextualised recommendation systems; ant colony optimisation; ACO; clustering algorithms; Maslow; hierarchy of needs; e-commerce; electronic commerce; customer behaviour; recommender systems; customer purchasing; metaheuristics; swarm intelligence; data sparseness; nearest neighbour search; collaborative filtering.
Int. J. of Wireless and Mobile Computing, 2016 Vol.11, No.4, pp.309 - 317
Submission date: 16 Jul 2016
Date of acceptance: 30 Nov 2016
Available online: 14 Feb 2017