Lise's INQuisitive Students
Machine Learning Research Group @ UMD
Statistical Relational Learning
Below are a fledgling collections of resources. Please email me if you have additions (or deletions!) that you would like to make to the lists.
Statistical Relational Learning (SRL) is an emerging area of machine learning research which attempts to combine the rich knowledge representation languages with statistical models, most often hierarchical Bayesian models. A number of approaches have been proposed, including ones that I've been quite involved in such as Probabilistic Relational Models from Daphne Koller's research group.
David Jensen and I organized a series of workshops in an effort to bring together researchers from diverse backgrounds that were interested in applying statistical modeling techniques to structured data:
- AAAI-2000 workshop on Learning Statistical Models from Relational Data (SRL2000)
- IJCAI 2003 workshop on Learning Statistical Models from Relational Data (SRL2003)
And, because the need to build probabilistic models of structure data occurs in so many sub-discplines such as computer vision, natural language processing, and social network analysis, to name a few, in 2004, Tom Dietterich, Kevin Murphy and I organized a workshop trying to bring all these folks together as well:
- ICML-2004 workshop, Statistical Relational Learning and its Connections to Other Fields (SRL2004)
The most recent event on this topic was an excellent Dagstuhl workshop, organized by Luc De Raedt with help from Tom Dietterich, Stephen Muggleton, and myself:
- Dagstuhl 2005 workshop on Probabilistic, Logical and Relational Learning - Towards a Synthesis
Avi Pfeffer and I will be giving a tutorial on Representation, Reasoning and Learning in Relational Probabilistic Languages at IJCAI 2005.
I will be giving an invited tutorial on Statistical Relational Learning at ILP 2005/ICML 2005.
Ben Taskar and I are currently editing a book on Statistical Relational Learning.