Enhancing Recommender Systems with Rich Side Information

PhD Thesis Proposal Defence


Title: "Enhancing Recommender Systems with Rich Side Information"

by

Mr. Huan ZHAO


Abstract:

Collaborative filtering (CF) based methods have become the most popular 
technique for recommender systems (RSs). In recent years, various types of 
side information such as social connections among users and metadata of 
items have been introduced into CF and shown to be effective for improving 
recommendation performance. Moreover, side information helps to alleviate 
data sparsity and cold start problems of conventional CF. However, 
previous works process different types of information separately, thus 
losing information that might exist across different types of side 
information.

In this proposal, we explore methods to enhance RS with various side 
information. We start with the incorporation of one important type of side 
information, i.e., social connections among users, into a state-of-the-art 
matrix factorization (MF) method, Local LOw Rank Matrix Approximation 
(LLORMA), and propose our Social LOcal Matrix Approximation (SLOMA). 
Experimental results on two real-world datasets demonstrate the 
superiority of SLOMA to LLORMA in the rating prediction task.

Next, we study the application of Heterogeneous Information Network (HIN), 
which offers a flexible representation for different types of information, 
to enhance CF based recommendation methods. Since HIN could be a complex 
graph representing multiple types of relations existing between two entity 
types, we need to tackle two challenging issues facing HIN-based RSs: How 
to capture the complex semantics driving the similarities between users 
and items in a HIN, and how to fuse the heterogeneous information to 
support recommendation. We propose to apply metagraph to HIN-based RSs to 
overcome the former problem and the ``matrix factorization (MF) + 
factorization machine (FM)'' framework for the latter. For the MF part, we 
obtain the user-item similarity matrix from each metagraph and then apply 
low-rank matrix approximation to obtain latent features for both users and 
items.  For the FM part, we propose to apply FM with Group lasso (FMG) on 
the features obtained from the MF part to train the recommendation model 
and at the same time identify the useful metagraphs. Experimental results 
on two large real-world datasets, i.e., Amazon and Yelp, show that our 
proposed approach is better than FM and other state-of-the-art HIN-based 
recommendation methods.

Finally, besides metagraph, we further propose Motif Enhanced MetaPath 
(MEMP) based similarities between users and items in HIN-based RSs. Motif 
is a local structure that can capture higher-order relations among nodes 
in homogeneous graphs. We argue that such higher-order relations also 
exist among nodes of same types in HIN. Thus, existing metapath based 
similarities can also be enhanced by integrating these motif-based 
higher-order relations. After computing the MEMP based similarities 
between users and items, our proposed ``MF+FM'' framework is adopted to 
fuse the similarities and for rating prediction. Experiments have been 
conducted on two real-word datasets, Epinions and CiaoDVD, the results 
demonstrate the effectiveness of MEMP in HIN-based RSs.


Date:			Monday, 22 October 2018

Time:                  	3:30pm - 5:30pm

Venue:                  Room 2408
                         (lifts 17/18)

Committee Members:	Prof. Dik-Lun Lee (Supervisor)
 			Dr. Yangqiu Song (Chairperson)
 			Dr. Wilfred Ng
 			Prof. Nevin Zhang


**** ALL are Welcome ****