The purpose of the Statistical Relational AI (StarAI) workshop is to bring together researchers and practitioners from two fields: logical (or relational) AI and probabilistic (or statistical) AI. These fields share many key features and often solve similar problems and tasks. Until recently, however, research in them has progressed independently with little or no interaction. The fields often use different terminology for the same concepts and, as a result, keeping-up and understanding the results in the other field is cumbersome, thus slowing down research.
Our long term goal is to change this by achieving a synergy between logical and statistical AI.
As a stepping stone towards realizing this big picture view on AI, we are organizing the Fourth International Workshop on Statistical Relational AI at the Twenty-Eighth AAAI Conference on Artificial Intelligence in Québec City, Québec, Canada, on July 27th 2014.
StarAI will be a one day workshop with around 50 attendees, paper presentations and poster spotlights, a poster session, and three invited speakers:
- Luc De Raedt (KU Leuven, Belgium)
- Vibhav Gogate (UT Dallas, USA)
- Henry Kautz (University of Rochester, USA)
Those interested in attending should submit either a technical paper (AAAI style, 6 pages without references) or a position statement (AAAI style, 2 pages maximum) in PDF format via EasyChair. All submitted papers will be carefully peer-reviewed by multiple reviewers and low-quality or off-topic papers will be rejected. Papers will be selected either for a short oral presentation or a poster presentation.
Paper Submission: April 17
Notification of Acceptance: May 1
Camera-Ready Papers: May 15
- Date of Workshop: July 27
- 8:50 a.m.: Welcome and introduction
- 9:00 a.m.: Invited talk by Luc De Raedt
Title: Progress in probabilistic logic programming
Probabilistic logic programs combine the power of a programming language with a possible world semantics, typically based on Sato's distribution semantics and they have been studied for over twenty years. In this talk, I shall report on recent progress within this paradigm. It will concern an extension towards dealing with continuous distributions as well as coping with dynamics. This is the framework of distributional clauses that has been applied to several applications in robotics, for tracking relational worlds in which objects or their properties are occluded in real time. I shall also report on an upgrade of the traditional rule learning paradigm to a probabilistic logical setting. It provides a novel perspective on learning the structure of SRL models in that it has traditional rule learners such as FOIL as a special case. The ProbFOIL system has been applied to learn rules in a probabilistic database setting and on data from CMU's Never Ending Language Learner.
Finally, I shall discuss some of the open challenges within probabilistic logic programming.
- 10:00 a.m.: Poster spotlights for papers 1 to 9
- 10:30 a.m.: Coffee break
- 11:00 a.m.: Invited talk by Vibhav Gogate
Title: Fast, Lifted, Sampling-Based Inference in Statistical Relational Models
Scaling up inference algorithms so that they can handle problems having millions of random variables and billions of features is currently the main challenge in statistical relational learning. In this talk, I will present my group's work on scaling up sampling or simulation based approximate inference algorithms to address this challenge. The talk will be structured in two parts. In the first part, I will present our work on exact lifted approaches, namely approaches that exploit exact symmetries and show how they can benefit importance sampling, Gibbs sampling and their numerous advanced variations. In the second part, I will present approaches that take advantage of approximate symmetries by leveraging Database theory, approximate model counting techniques and clustering algorithms. I will conclude my presentation by discussing avenues for future work and open problems.
- 12:00 p.m.: Poster spotlights for papers 10 to 18
- 12:30 p.m.: Lunch break
- 2:00 p.m.: Invited talk by Henry Kautz
Title: A Markov Logic Framework for Recognizing Complex Events from Multimodal Data
We present a general framework for complex event recognition that is well-suited for integrating information that varies widely in detail and granularity. Consider the scenario of an agent in an instrumented space performing a complex task while describing what he is doing in a natural manner. The system takes in a variety of information, including objects and gestures recognized by RGB-D and descriptions of events extracted from recognized and parsed speech. The system outputs a complete reconstruction of the agent's plan, explaining actions in terms of more complex activities and filling in unobserved but necessary events. We show how to use Markov Logic (a probabilistic extension of first-order logic) to create a model in which observations can be partial, noisy, and refer to future or temporally ambiguous events; complex events are composed from simpler events in a manner that exposes their structure for inference and learning; and uncertainty is handled in a sound probabilistic manner. We demonstrate the effectiveness of the approach for tracking kitchen activities in the presence of noisy and incomplete observations.
- 3:00 p.m.: Poster spotlights for papers 19 to 27
- 3:30 p.m.: Poster session (with coffee)
- 5:00 p.m.: End
The papers are available for download on the AAAI workshop proceedings website.
- Udi Apsel, Kristian Kersting and Martin Mladenov.
Tightening Relational MAP-LPs via Symmetry of Clusters
- Zhengya Sun, Zhuoyu Wei and Jue Wang.
Scalable Learning for Structure in Markov Logic Networks
- Brian Ruttenberg, Matthew Wilkins and Avi Pfeffer.
Hierarchical Reasoning with Probabilistic Programming
- Golnoosh Farnadi, Stephen H. Bach, Marie-Francine Moens, Lise Getoor and Martine De Cock.
Extending PSL with Fuzzy Quantifiers
- William Yang Wang, Kathryn Mazaitis and William W Cohen.
ProPPR: Efficient First-Order Probabilistic Logic Programming for Structure Discovery, Parameter Learning, and Scalable Inference
- Joris Renkens, Angelika Kimmig, Guy Van den Broeck and Luc De Raedt.
Explanation-Based Approximate Weighted Model Counting for Probabilistic Logics
- Ismail Ilkan Ceylan and Rafael Peñaloza.
Reasoning in the Description Logic BEL using Bayesian Networks
- Fatemeh Riahi and Oliver Schulte.
A Proposal for Statistical Outlier Detection in Relational Structures
- Islam Beltagy and Raymond Mooney.
Efficient Markov Logic Inference for Natural Language Semantics
- Seyed Mehran Kazemi, David Buchman, Kristian Kersting, Sriraam Natarajan and David Poole.
Relational Logistic Regression: the Directed Analog of Markov Logic Networks
- Tushar Khot, Sriraam Natarajan and Jude Shavlik.
Classification from one class of examples for relational domains
- Aniruddh Nath and Pedro Domingos.
Learning Tractable Statistical Relational Models
- Aniruddh Nath and Pedro Domingos.
Automated Debugging with Tractable Probabilistic Programming
- Masaaki Nishino, Akihiro Yamamoto and Masaaki Nagata.
A Sparse Parameter Learning Algorithm for Probabilistic Logic Programs
- Shrutika Poyrekar, Sriraam Natarajan and Kristian Kersting.
A Deeper Empirical Analysis of CBP algorithm: Grounding is the Bottleneck
- Jonas Vlasselaer, Wannes Meert, Guy Van den Broeck and Luc De Raedt.
Efficient Probabilistic Inference for Dynamic Relational Models
- Mihaela Verman, Philip Stutz and Abraham Bernstein.
Solving Distributed Constraint Optimization Problems using Ranks
- Eric Gribkoff, Guy Van den Broeck and Dan Suciu.
Understanding the Complexity of Lifted Inference and Asymmetric Weighted Model Counting
- Deepak Venugopal and Vibhav Gogate.
Evidence-based Clustering for Scalable Inference in Markov Logic
- Jaesik Choi, Eyal Amir, Tianfang Xu and Albert Valocchi.
Parameter Estimation for Relational Kalman Filtering
- Matthew Dirks, Andrew Csinger, Andrew Bamber and David Poole.
Representation, Reasoning and Inference for a Relational Open-World Influence Diagram Applied to a Real-Time Geological Domain
- Sebastian Riedel, Sameer Singh, Vivek Srikumar, Tim Rocktaschel, Larysa Visengeriyeva and Jan Noessner.
Wolfe: Strength Reduction and Approximate Programming for Probabilistic Programming
- Junkyu Lee, Radu Marinescu and Rina Dechter.
Applying Marginal MAP Search to Probabilistic Conformant Planning
- Mathias Niepert and Pedro Domingos.
Tractable Probabilistic Knowledge Bases: Wikipedia and Beyond
- Mathias Niepert and Guy Van den Broeck.
Tractability through Exchangeability: A New Perspective on Efficient Lifted Inference
- Tim Rocktaschel, Matko Bošnjak, Sameer Singh and Sebastian Riedel.
Low-Dimensional Embeddings of Logic
- Daniel Lowd, Brenton Lessley and Mino De Raj Manoharan.
Towards Adversarial Reasoning in Statistical Relational Domains
For comments, queries and suggestions, please contact:
- Guy Van den Broeck
(UCLA, KU Leuven)
- Kristian Kersting
(TU Dortmund, Fraunhofer IAIS)
- Sriraam Natarajan
- David Poole
(University of British Columbia)
- Hendrik Blockeel (KU Leuven)
- Rodrigo Braz (SRI International)
- Arthur Choi (UCLA)
- Jesse Davis (KU Leuven)
- Luc De Raedt (KU Leuven)
- Pedro Domingos (University of Washington)
- Paolo Frasconi (Università degli Studi di Firenze)
- Vibhav Gogate (UT Dallas)
- Tuyen Huynh (SRI International)
- Manfred Jaeger (Aalborg University)
- Roni Khardon (Tufts University)
- Tushar Khot (University Of Wisconsin-Madison)
- Angelika Kimmig (KU Leuven)
- Niels Landwehr (University of Potsdam)
- Daniel Lowd (University of Oregon)
- Brian Milch (Google)
- Kee Siong Ng (EMC Greenplum)
- Mathias Niepert (University of Washington)
- Scott Sanner (NICTA)
- Vítor Santos Costa (Universidade do Porto)
- Taisuke Sato (Tokyo Institute of Technology)
- Jude Shavlik (University of Wisconsin-Madison)
- Sameer Singh (University of Massachusetts, Amherst)
StarAI is currently provoking a lot of new research and has tremendous theoretical and practical implications.
Theoretically, combining logic and probability in a unified representation and building general-purpose reasoning tools for it has been the dream of AI, dating back to the late 1980s.
Practically, successful StarAI tools will enable new applications in several large, complex real-world domains including those involving big data, social networks, natural language processing, bioinformatics, the web, robotics and computer vision. Such domains are often characterized by rich relational structure and large amounts of uncertainty. Logic helps to effectively handle the former while probability helps her effectively manage the latter.
We seek to invite researchers in all subfields of AI to attend the workshop and to explore together how to reach the goals imagined by the early AI pioneers.
The focus of the workshop will be on general-purpose representation, reasoning and learning tools for StarAI as well as practical applications. Specifically, the workshop will encourage active participation from researchers in the following communities: satisfiability (SAT), knowledge representation (KR), constraint satisfaction and programming (CP), (inductive) logic programming (LP and ILP), graphical models and probabilistic reasoning (UAI), statistical learning (NIPS and ICML), graph mining (KDD and ECML PKDD) and probabilistic databases (VLDB and SIGMOD).
It will also actively involve researchers from more applied communities, such as natural language processing (ACL and EMNLP), information retrieval (SIGIR, WWW and WSDM), vision (CVPR and ICCV), semantic web (ISWC and ESWC) and robotics (RSS and ICRA).
Previous StarAI workshops were held in conjunction with AAAI 2010, UAI 2012, and AAAI 2013, and were among the most popular workshops at the conferences.