Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance...

35
Collaborative Filtering with Entity Similarity Regularization in Heterogeneous Information Networks Xiao Yu 1 , Xiang Ren 1* , Quanquan Gu 1 , Yizhou Sun 2 , Jiawei Han 1 1 Univ. of Illinois, at Urbana-Champaign 2 Northeastern Univ. *[email protected] 1

Transcript of Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance...

Page 1: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Co l l a bo rat ive F i l te r i ng wi t h Entity Similarity Regularization in Heterogeneous Information Networks

Xiao Yu1, Xiang Ren1*, Quanquan Gu1, Yizhou Sun2, Jiawei Han1

1Univ. of Illinois, at Urbana-Champaign 2Northeastern Univ. *[email protected]

1

Page 2: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Roadmap

• Why Study CF in HIN?

• Background and Preliminaries

• Proposed Method

• Experiments

• Conclusion and Future Work

2

Page 3: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Recommender Systems are Everywhere!

3

Page 4: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Recommendation Paradigm

4

user profiles

I1 I2 … Im

U1 ? ? ? 5

U2 ? 3 ? 4

… ? ? ? ?

Un 2 1 ? ?

user item ratings

item features

external knowledge

recommender system recommendation

Page 5: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Recommender System with Network

• Utilizing network relationship information can enhance the recommendation quality

• However, most of the previous studies only use single type of relationship between users or items (e.g., social network [Ma,WSDM11], trust relationship [Ester, KDD10], service membership [Yuan, RecSys11])

5

Page 6: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

The Heterogeneous Information Network View of Recommender System

6

Page 7: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Why Information Network Can Help?

• Various types of information and relationships complement each other.

• Number of ratings - power law distribution

• Cold Start – How to handle new users or new items?

7

# of ratings

A very small number of users and items have a lot of ratings

Most users and items do not have enough ratings

nu

ms

of

use

rs

Page 8: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Roadmap

• Why Study CF in HIN?

• Background and Preliminaries

• Proposed Method

• Experiments

• Conclusion and Future Work

8

Page 9: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

What Are Information Networks? • A network where each node represents an entity (e.g.,

user in a social network) and each link (e.g., friendship)

a relationship between entities.

– Nodes/links may have attributes, labels, and weights.

– Links may carry rich semantic information.

9

Page 10: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Heterogeneous Information Networks

10

Venue Paper Author

DBLP Bibliographic Network The IMDb Movie Network

Actor

Movie

Director

Movie

Studio

The Facebook Network

1. Multiple entity types and link types 2. New problems are emerging in heterogeneous networks!

Page 11: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Heterogeneous Information Networks Are Ubiquitous

11

Social Media Protein Networks E-commerce

Medical

Database Medical

Images

Medical

Records

Treatment Plan

Pharmacy Service

Healthcare Knowledge Graph

Page 12: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

IMDb Network Schema

12

background

Page 13: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Entity Similarity

13

In heterogeneous information networks, find entities which are similar to a given entity query.

In DBLP, who are similar to “C. Faloutsos”?

In IMDb, which TVs / movies are similar to “Avatar”?

In Yelp, which restaurants are similar to “Blackdog”?

background

Page 14: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Meta-Path [Sun, VLDB 2011]

14

A1

A2

P1

P2

VLDB

Social Network

A3

A4

Network Snippet

• Meta-level description of a path between two entities • A path on network schema • Denote an existing or concatenated relation between two

entity types

A1-P1-A2 A1-P1-VLDB-P3-A3 A1-P1-”Social Network”-P2-A4 ……

P3

A1 is similar to A2, A3 and A4 but why?

Author-Paper-Author Author-Paper-Venue-Paper-Author Author-Paper-Term-Paper-Author

background

Page 15: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Similarity Measurement

• PathSim [Sun, VLDB 2011]

• Normalized path count between x and y following meta-path 𝒫

• Entities with strong connectivity and similar visibility under the given meta-path

– Path Constrained Random Walk[Lao, Machine Learning, 2010]

15

Visibility of x Visibility of y

background

Page 16: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Different Meta-Paths Carry Different Semantics

• Who are most similar to C. Faloutsos?

16

Christos’s students or close collaborators Work on similar topics and have similar reputation

Meta-Path: Author-Paper-Author Meta-Path: Author-Paper-Venue-Paper-Author

background

Page 17: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Problem Definition

• Given

• For a specific user, find items of interests based his / her previous rating history.

17

E1 e2 … em

u1 0 0 0 1

u2 0 2 0 5

… 0 0 0 0

un 3 4 0 0

Rating Data Information Network

Page 18: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Roadmap

• Why Study CF in HIN?

• Background and Preliminaries

• Proposed Method

• Experiments

• Conclusion and Future Work

18

Page 19: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Notations

• We have n users and m items.

• By computing similarity scores of all item pairs along certain meta-pat, we can get a similarity matrix.

• With L different meta-paths, we can calculate L similarity matrices as

19

Page 20: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Traditional Matrix Factorization

• Approximate R with product of U and V

• Non-Negative Matrix Factorization

• Weighted Non-Negative matrix Factorization

20

Page 21: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Objective Function

21

Approximate R with U V product Regularization on U V

Regularization on θ Similar items measured from HIN should have similar low-rank representations

Page 22: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Simplify Optimization Process

22

where

Page 23: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Revised Objective Function

23

Similar items measured from HIN should have similar low-rank representations

Page 24: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Parameter Estimation

24

Step 1

Step 2

Step 3

Iteratively updating U, V and θ till convergence

Page 25: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Roadmap

• Why Study CF in HIN?

• Background and Preliminaries

• Proposed Method

• Experiments

• Conclusion and Future Work

25

Page 26: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Dataset • We combine IMDb + MovieLens100K

26

We random sample training datasets of different sizes (0.4, 0.6, and 0.8)

Page 27: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Comparison Methods

27

We use Hete-MF to represent the proposed method.

Page 28: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Evaluation Metrics

• We use Mean Absolute Error and Root Mean Square Error to evaluate the performance.

28

Page 29: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Performance Comparison

29

Page 30: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Performance Analysis

30

Page 31: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Convergence Rate

31

Page 32: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Roadmap

• Why Study CF in HIN?

• Background and Preliminaries

• Proposed Method

• Experiments

• Conclusion and Future Work

32

Page 33: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Conclusions

• We study CF in HIN.

• We combine rating data with meta-path-based similarity matrices.

• We compared the proposed approaches with several widely employed or state-of-the-art recommendation techniques.

• We analyzed the performance of these methods under different scenarios.

33

Page 34: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Future Work

• Adding user and/or item rating priors to the proposed method to alleviate cold start problem

• Personalized recommendation models

• On-line version of the method to incorporate newly generated ratings

34

Page 35: Collaborative Filtering with Entity Similarity …ink-ron.usc.edu/xiangren/ijcai13_HINA.pdfenhance the recommendation quality •However, most of the previous studies only use single

Thank You!!

35