Extended Deadline: 8th March

1st International Workshop on Formalisms and Methodology for Learning by Reading (FAM-LbR)
NAACL 2010 Workshop
June 5-6, 2010
(http://www.rutumulkar.com/FAM-LbR.php )

Call for Papers

It has been a long term vision of Artificial Intelligence to develop Learning by Reading systems that can capture knowledge from naturally occurring texts, convert it into a deep logical notation and perform some inferences/reasoning on them. Such systems directly build on relatively mature areas of research, including Information Extraction (for picking out relevant information from the text), Commonsense and AI Reasoning (for deriving inferences from the knowledge acquired), Bootstrapped Learning (for using the learned knowledge to expand the knowledge base) and Question Answering (for providing evaluation mechanisms for Learning by Reading systems). In Natural Language Processing, statistical learning techniques have provided new solutions and breakthroughs in various areas over the last decade. In Knowledge Representation and Reasoning, systems have achieved impressive performance and scale in far more complex problems than the past.

Learning by Reading is a two-part process. One part deals with extracting interesting information from naturally occurring texts, and the other is to use this extracted knowledge to expand the knowledge base and consequently the system's inference capabilities. Previous systems have chosen either a "broad and shallow" or a "narrow and deep" knowledge acquisition and reasoning strategy. These techniques are constrained by either their limited reasoning ability or their extreme domain dependence.

The goal of this workshop is to draw together researchers to explore the nature and degree of integration possible between symbolic and statistical techniques for knowledge acquisition and reasoning. In particular, given these developments, what is the role of commonsense knowledge and reasoning in language understanding? What are the limitations of each style of processing, and how can they be overcome by complementary strengths of the other? What are appropriate evaluation metrics for Learning by Reading systems?


Topics of interest include (but are not limited to)
------------------------------ ----------------------
Unguided and targeted (goal-directed) machine reading
Wikipedia and web based machine reading
Knowledge extraction from text vs. using pre-built knowledge resources
Learning temporal sequences, causality, and other semantics from text
Bridging knowledge gaps in text through inference
Ontology learning or expansion
Knowledge Integration into evolving models
Abductive/deductive, commonsense, and other reasoning
Bootstrapping learning by Reading systems


Important Dates
-----------------
Mar 8, 2010 Submission due date
Mar 30, 2010 Notification of acceptance
Apr 12, 2010 Camera ready papers due
Jun 5-6, 2010 Workshops


Submission Instructions
------------------------
Please visit http://www.rutumulkar.com/FAM-LbR.php for more information.


Location
---------
FAM-LbR is held with NAACL 2010 (June 1-6, 2010) in downtown Los Angeles. Local information can be found from the conference website (http://naaclhlt2010.isi.edu/index.html).


Related Workshops and Conferences
------------------------------ ----
Machine Reading, AAAI Spring Symposium 2007 (http://www.cs.washington.edu/homes/pjallen/aaaiss07/index.htm)
Learning by Reading and Learning to Read, AAAI Spring Symposium 2009 (http://www.coral-lab.org/~oates/aaai2009ss/)
K-CAP 2009 (http://kcap09.stanford.edu/)


Organizers
-----------
Rutu Mulkar-Mehta
James Allen
Jerry Hobbs
Eduard Hovy
Bernardo Magnini
Chris Manning

Contact Information
--------------------
Please email Rutu Mulkar-Mehta (me@rutumulkar.com) for any further questions.