Automated Dimension Determination for NMF-based Incremental Collaborative Filtering
Journal Title: EAI Endorsed Transactions on Collaborative Computing - Year 2015, Vol 1, Issue 5
Abstract
The nonnegative matrix factorization (NMF) based collaborative filtering t e chniques h a ve a c hieved great success in product recommendations. It is well known that in NMF, the dimensions of the factor matrices have to be determined in advance. Moreover, data is growing fast; thus in some cases, the dimensions need to be changed to reduce the approximation error. The recommender systems should be capable of updating new data in a timely manner without sacrificing the prediction accuracy. In this paper, we propose an NMF based data update approach with automated dimension determination for collaborative filtering purposes. The approach can determine the dimensions of the factor matrices and update them automatically. It exploits the nearest neighborhood based clustering algorithm to cluster users and items according to their auxiliary information, and uses the clusters as the constraints in NMF. The dimensions of the factor matrices are associated with the cluster quantities. When new data becomes available, the incremental clustering algorithm determines whether to increase the number of clusters or merge the existing clusters. Experiments on three different datasets (MovieLens, Sushi, and LibimSeTi) were conducted to examine the proposed approach. The results show that our approach can update the data quickly and provide encouraging prediction accuracy.
Authors and Affiliations
Xiwei Wang, Jun Zhang, Ruxin Dai
Keywords
Related Articles
- EP ID EP45700
- DOI http://dx.doi.org/10.4108/eai.17-12-2015.150804
- Views 313
- Downloads 0
A Novel, Privacy Preserving, Architecture for Online Social Networks
The centralized nature of conventional OSNs poses serious risks to the privacy and security of information exchanged between their members. These risks prompted several attempts to create decentralized OSNs, or DOSNs. Th...
Matching with Stochastic Arrival
We study matching in a dynamic setting, with applications to the allocation of public housing. In our model, objects of different types that arrive stochastically over time must be allocated to agents in a queue. For the...
An Analytical Study of Computation and Communication Tradeoffs in Distributed Graph
Distributed vertex-centric graph processing systems such as Pregel, Giraph and GPS have acquired significant popularity in recent years. Although the manner in which graph data is partitioned and placed on the computatio...
MOSDEN: A Scalable Mobile Collaborative Platform for Opportunistic Sensing Applications
Mobile smartphones along with embedded sensors have become an efficient enabler for various mobile applications including opportunistic sensing. The hi-tech advances in smartphones are opening up a world of possibilities...
A Multimodal Dataset for the Analysis of Movement Qualities in Karate Martial Art
A multimodal dataset is presented, which has been collected for analyzing and measuring the quality of movement performed during sport activities. Martial arts (namely karate) are taken as test-beds for investigation. Ka...