IS 2013 - IEEE Intelligent Systems Special Issue on Transfer Learning in Web and Social-Network Mining

Website | Edit Freely

Category IS 2013

Deadline: September 01, 2012 | Date: June 01, 2013

Venue/Country: Online, Online

Updated: 2012-02-09 14:52:51 (GMT+9)

Call For Papers - CFP

IEEE Intelligent Systems

Special Issue on Transfer Learning in Web and Social-Network Mining

Publication: May/June 2013

Submissions due for review: 1 September 2012

In the field of information retrieval and Web mining, more and more

learning tasks can easily acquire multiple datasets from various

domains. For example, many of today's recommendation tasks are

starting to leverage multiple types of user data from different

domains, such as users' browsing-history data, shopping-record data,

and social-network (SN) data. At the same time, the need for knowledge

transfer is increasingly evident as many new datasets, or parts of

data, are only very sparsely annotated.

Different from traditional single-domain learning problems (which are

based on the assumption that training and test data are drawn from

identical distributions), transfer learning problems are built on

multiple-domain data that may have different degrees of relatedness to

target tasks. This offers an opportunity for different related

applications to help one another acquire knowledge. To better leverage

multiple-domain data, the mining and transferring of shared knowledge

across multiple domains is likely to become a crucial step in

information retrieval (IR), recommendation, and Web and SN mining in

the future.

This special issue is dedicated to transfer learning for IR and Web

mining. It will bring together researchers from information retrieval

and machine learning, collaborative recommendations, natural language

processing (NLP), social networks, and other areas of computer and

information science who are working on or interested in this area.

There are many challenging research issues to explore. They fallmainly

into two categories:

+ general transfer-learning methods in various multi-domain data types

that arise in IR, recommendation, and Web mining (examples include

heterogeneous, structured, or stream data types), and their related

foundations and theories; and

+ specific transfer-learning methods for various important IR,

collaborative filtering (CF), and Web mining tasks (examples include

transfer learning for ranking, recommendation, NLP, social network

analysis, user profiling, micro-blogging applications, and information

extraction, as well as other novel applications built on multiple

domain data).

This issue will provide a forum for these researchers to identify

issues and challenges, share their latest results, express a diverse

range of opinions about this topic, and discuss future directions. We

believe this issue will become an important milestone in the

development of this new area of IR and Web and SN mining.

*Submission Guidelines*

Submissions should not only describe technical research but also show

its benefits in terms of sustainability. A special emphasis will be

given to research that has matured beyond the design, laboratory, or

simulation stage and that reports experiences and lessons learned from

deployment in the field.

As distributed AI for sustainability involves (often rather

disruptive) sociotechnical innovations, we also welcome

interdisciplinary consideration of social and economic aspects, such

as new sustainable business models, customer service value

propositions, market innovations, incentives for adoption,

interoperability and standardization issues, and regulatory aspects.

Submissions should be 3,000 to 5,400 words (counting a standard figure

or table as 200 words) and should follow IEEE Intelligent Systems

style and presentation guidelines (www.computer.org/intelligent/

author). The manuscripts cannot have been published or be currently

submitted for publication elsewhere.

We strongly encourage submissions that include audio, video, and

community content, which will be featured on the IEEE Computer Society

Web site along with the accepted papers.

*Further Resources*

+ Recent publications on the topic of transfer learning at top

conferences: http://www1.i2r.a-star.edu.sg/~jspan/conferenceTL.htm

+ "A Transfer Learning Survey" by Sinno Jialin Pan and Qiang Yang,

IEEE Transactions on Knowledge and Data Engineering, vol. 22, no. 10,

Oct. 2010; http://www1.i2r.a-star.edu.sg/~jspan/publications/TLsurvey_0822.pdf.

+ "Transfer Learning Resources," including dataset and code

repositories; www.cse.ust.hk/TL/index.html

+ Recent workshops and symposia:

- "Workshop on Unsupervised and Transfer Learning";

http://clopinet.com/isabelle/Projects/ICML2011

- 25th Annual ConferenceonNeural Information Processing Systems;

http://nips.cc/Conferences/2009/Program/event.php?ID=1527

*Guest Editors*

+ Deepak Agarwal, Yahoo! Labs

+ Karsten Borgwardt, Max Planck Insts. & Eberhard Karls Univ.

Tuebingen

+ Yi Chang, Yahoo! Labs

+ Bo Long, Yahoo! Labs

+ Qiang Yang, Hong Kong University of Science and Technology

*Questions?*

+ Information about the special issue's focus: contact Qiang Yang at

qyangcse.ust.hk (include the keyword "IEEE IS: Transfer Learning" in

the subject line)

+ General author guidelines: www.computer.org/intelligent/author

+ Submission details: contact intelligentcomputer.org

+ To submit an article: https://mc.manuscriptcentral.com/is-cs (log

in, then select "Special Issue on Transfer Learning")