=Paper=
{{Paper
|id=Vol-1887/preface
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1887/preface.pdf
|volume=Vol-1887
}}
==None==
https://ceur-ws.org/Vol-1887/preface.pdf
The 1st Workshop on Intelligent Recommender Systems by
Knowledge Transfer & Learning (RecSysKTL)
Yong Zheng Weike Pan
School of Applied Technology College of Computer Science & Software Engineering
Illinois Institute of Technology Shenzhen University
Chicago, Illinois, USA 60616 Shenzhen, Guangdong, China 518060
yzheng66@iit.edu panweike@szu.edu.cn
Shaghayegh (Sherry) Sahebi Ignacio Fernández
Department of Computer Science NTENT
University at Albany, SUNY Almogàvers 123
Albany, NY, USA 12222 Barcelona, Spain 08018
ssahebi@albany.edu ignacio.fernandezt@uam.es
ABSTRACT viewpoints to provide better-quality recommendations and resolve
Cross-domain recommender systems and transfer learning approach- issues such as the cold-start problem.
es are useful to help integrate knowledge from different places, so The emerging research on cross-domain, context-aware and
that we alleviate some existing problems (such as the cold-start multi-criteria recommender systems, has proved to be successful.
problem), or improve the quality of recommender systems. With Given the recent availability of cross-domain datasets and novelty
the advantages of these techniques, we host the first international of the topic, we organize the 1st workshop on intelligent recom-
workshop on intelligent recommender systems by knowledge trans- mender systems by knowledge transfer and learning (RecSysKTL)
fer and learning (RecSysKTL) to provide such a forum for academia held in conjunction with the 11th ACM Conference on Recom-
researchers and application developers from around the world to mender Systems. The definition of “domain" may vary in different
present their work and discuss exciting research ideas or outcomes. applications, e.g., it could be (but not limited to):
The workshop is held in conjunction with the ACM Conference on
Recommender Systems 2017 on August 27th at Como, Italy. • From one application to another: We may utilize user be-
haviors on social networks to predict their preferences on
CCS CONCEPTS movies (e.g., Netflix, Youtube) or music (e.g., Pandora, Spotify).
• From one category to another: We may predict a userŠs
• Information systems → Recommender systems;
taste on electronics by using his or her preference history on
books based on the data collected from Amazon.com.
KEYWORDS • From one context to another: We may collect a userŠs pref-
cross-domain, knowledge transfer, recommender system erences on the items over different time segment (e.g., weekend
or weekday) and predict her preferences on movie watching
within another context (e.g., companion and location).
1 BACKGROUND AND MOTIVATIONS • From one task to another: It may be useful for us to predict
Recommender systems, as one of well-known Web intelligence how a user will select hotels for his or her vocations by learning
applications, aim to alleviate the information overload problem and from how he or she books the tickets for transportation.
produce item suggestions tailored to user preferences. Typically, • From one structure to another: It could be also possible for
user preferences or tastes are collected through usersŠ implicit or us to infer social connections by learning from the structure of
explicit feedback in various formats, such as user ratings, online heterogeneous information network.
behaviors, text reviews, etc. Also, user feedback on different items
can be collected from several systems or domains. The diversity of
feedback formats and domains provides multiple views to usersŠ 2 OBJECTIVES
preferences, and thus, can be helpful in recommending more related This workshop intends to create a medium to generate more practi-
items to users. Cross-domain recommender systems and transfer cal and efficient predictive models or recommendation approaches
learning approaches propose to take advantage of such diversity of by leveraging user feedbacks or preferences from multiple domains.
This workshop will be beneficial for both researchers in academia
RecSysKTL Workshop @ ACM RecSys ’17, August 27, 2017, Como, Italy and data scientists in industry to explore and discuss different
© 2017 Copyright is held by the author(s).
definition of domains, interesting applications, novel predictive
models or recommendation approaches to serve the knowledge
transfer and learning from one domain to another.
As this is the first time to host such a workshop on knowledge
transfer and learning in the area of recommender systems, we also
expect to collect feedbacks from the workshop, so that we can better 4 PROGRAM COMMITTEES
organize the workshop in the future. We thank the contributions by the following workshop program
committee members:
3 TOPICS OF INTEREST • Alejandro Bellogín, Universidad Autónoma de Madrid
The topics of interest include (but are not limited to): • Steve Bourke, Schibsted Media Group
Applications of Knowledge Transfer for RecSys • Iván Cantador, Universidad Autónoma de Madrid
• Cross-domain recommendation • Liang Dong, Google, Inc.
• Context-aware or time-aware recommendation • Mehdi Elahi, Free University of Bozen
• Multi-criteria recommender systems • Negar Hariri, Apple, Inc.
• Novel applications • Mahesh Joshi, Linkedin
Methods for Knowledge Transfer in RecSys • Bin Li, Data61, Australia
• Knowledge transfer for content-based filtering • Zhongqi Lu, Hong Kong University of Science and Technology,
• Knowledge transfer in user- and item-based collaborative fil- Hong Kong, China
tering • Cataldo Musto, University of Bari "Aldo Moro"
• Transfer learning of model-based approaches to collaborative • Denis Parra, Pontificia Universidad Catolica de Chile
filtering • Alan Said, University of Skövde, Sweden
• Deep Learning methods for knowledge transfer • Yue Shi, Facebook, USA
• Fatemeh Vahedian, DePaul University
Challenges in Knowledge Transfer for Recommendation
• Saúl Vargas, Mendeley
• Addressing user feedback heterogeneity from multiple domains • Tong Yu, Carnegie Mellon University, USA
(e.g. implicit vs. explicit, binary vs. ratings, etc.) • Fuzhen Zhuang, Chinese Academy of Sciences, China
• Multi-domain and multi-task knowledge representation and • Yong Zhuang, Carnegie Mellon University, USA
learning
• Detecting and avoiding useless knowledge transfer 5 WORKSHOP PROGRAMS
• Ranking and selection of auxiliary sources of knowledge to
This year, we received 11 valid submissions. We finally accepted
transfer from
5 long and 2 short papers. The copyrights of the accepted papers
• Performance and scalability of knowledge transfer approaches
are held by the owner/author(s). All of the accepted papers will
for recommendation
be published by CEUR workshop proceedings1 . Hopefully, we can
Evaluation of RecSys based on Knowledge Transfer collaborative with some journals, and invite the extended version
• Beyond accuracy: novelty, diversity, and serendipity of recom- of the accepted papers to a special issue in the journal.
mendations supported by the transfer of knowledge
• Performance of knowledge transfer systems in cold-start sce- 6 ACKNOWLEDGMENTS
narios We thank the workshop organizing committee for giving us the
• Impact of the size and quality of transferred data on target opportunity to host this workshop in conjunction with ACM Rec-
recommendations Sys 2017. And we appreciate the contributions by the authors and
• Analysis of the amount of domain overlap on recommendation program committee members.
performance
1 http://ceur-ws.org/