the field of Web and social network mining, more and more learning
tasks can easily acquire multiple data sets from various domains. For
example, a modern search engine system often conducts ranking learning
tasks in various domains with different languages (e.g., English text
search, Spanish text search, etc.), or different verticals/topics
(e.g., news search, product search, etc.); and recently recommendation
systems start to leverage multiple types of user data from different
domains, such as user browsing history data, user shopping record data,
and user social network 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.
single-domain learning problems based on the assumption that training
and test data are drawn from identical distribution, cross domain
learning problems are built on multiple domain data that may have
different degrees of relatedness to target tasks, offering an
opportunity to help one another. To better leverage multiple domain
data, mining and transferring of shared knowledge across multiple
domains is likely to become a crucial step in Web and social network
mining in the future.
Topics of Interest
issues to be addressed include but are not limited to:
domain social network analysis
domain natural language processing
domain learning on structure data
domain learning on stream data
domain learning on heterogeneous data
Papers should be no more than 8 pages in length. The format should be the same as KDD main track format.
All submissions must be in PDG format and must no exceed 10MB in size.
Top quality papers will be invited to IEEE Inteligent System Special Issue.
paper submission site is here.
Submission deadline: May 12, 2012.