AAAI-2000 Workshop
Learning Statistical Models from Relational Data
July 31, 2000, Austin, TX
http://robotics.stanford.edu/srl

Researchers from a variety of backgrounds (including machine learning,
statistics, inductive logic programming, databases, and reasoning
under uncertainty) are beginning to develop techniques to learn
statistical models from relational data.  This work diverges from
traditional approaches in these fields that assume data instances are
structurally identical and statistically independent or assume that
relationships are deterministic.  New developments in this area are
vital because of the growing interest in mining information in
relational databases, object-oriented databases, XML and other
structured and semi-structured formats.  The workshop will focus on
learning models that represent statistical correlations between the
properties of related entities directly from relational data.

Central topics include:

o Methods for learning statistical models from heterogeneous,
   non-independent instances.

o Non-propositional data representations (including relational and
   first-order models).

o Efficient techniques for mining relational and semi-structured data.

o Applications of relational data analysis (e.g., Web mining,
   counter-terrorism, intrusion detection, collaborative filtering,
   bioinformatics).

Authors are invited to submit an extended abstract on the topics
outlined above.  Abstracts should emphasize technical research
results, either in the form of system capabilities or general
findings.  Abstracts should be no longer than 4 pages, and follow the
AAAI style sheet. Electronic submissions, in PostScript or PDF, should
be sent to srl-submit@robotics.stanford.edu. Accepted submissions will
be asked to submit a final version (up to 8 pages) of the paper and
may be asked to give an oral presentation at the workshop.  All papers
will be distributed and included in an AAAI Press technical report.