[UAI] ICAIL 2013 call for papers, workshop proposals and demonstrations

2012-11-22 Thread Bart Verheij
14th International Conference on
Artificial Intelligence & Law (ICAIL 2013)
June 10 – June 14, 2013
ITTIG-CNR
Consiglio Nazionale delle Ricerche (National Research Council of Italy)
Rome, Italy
http://icail2013.ittig.cnr.it


Sponsored by:
The International Association for Artificial Intelligence and Law (IAAIL)
ITTIG-CNR


Call for Papers, Workshop Proposals and Demonstrations

The field of AI and Law is concerned with:

* the study of legal reasoning using computational methods
* the study of AI and other advanced information technologies, using
law as an example domain
* formal models of norms, normative systems, norm-governed societies
* legal and quasi-legal applications of AI and other advanced
information technologies

The ICAIL conference is the primary international conference
addressing research in Artificial Intelligence and Law, and has been
organized biennially since 1987 under the auspices of the
International Association for Artificial Intelligence and Law (IAAIL).
ICAIL provides a forum for the presentation and discussion of the
latest research results and practical applications; it fosters
interdisciplinary and international collaboration. The conference
proceedings are published by ACM. The journal Artificial Intelligence
and Law regularly publishes expanded versions of selected ICAIL
papers.

ICAIL 2013, the fourteenth International Conference on Artificial
Intelligence and Law, invites the submission of papers on a broad
spectrum of research topics. Authors are invited to submit papers on
topics including but not restricted to

* Formal and computational models of legal reasoning
* Knowledge acquisition techniques for the legal domain, including
natural language processing and data mining
* Computational models of argumentation and decision making
* Legal knowledge representation including legal ontologies and common
sense knowledge
* Automatic legal text classification and summarization
* Automated information extraction from legal databases and texts
* Machine learning and data mining applied to legal databases
* E-discovery and e-disclosure
* E-government and e-justice
* Computational models of evidential reasoning
* Modeling norms for multi-agent systems
* Modeling negotiation and contract formation
* Computational models of case-based legal reasoning
* Conceptual or model-based legal information retrieval
* Online dispute resolution
* Intelligent legal tutoring systems
* Intelligent support systems for the legal domain
* Interdisciplinary applications of legal informatics methods and systems

Invited speakers

* Rosaria Conte, ISTC-CNR
* Paul Thagard, University of Waterloo
* Radboud Winkels, University of Amsterdam

Two tracks: regular papers and innovative applications papers

For ICAIL 2013, authors are invited to submit papers in one of two
tracks: regular and innovative applications.  In addition to papers
about results and findings from systems, approaches, or theoretical
models (in the conference's regular track), we encourage the
submission of original papers about innovative applications. Both
regular track papers and innovative applications papers will be
assessed in a rigorous reviewing procedure. Standard assessment
criteria for research papers will apply to all submissions (relevance,
originality, significance, technical quality, presentation). Papers
proposing formal or computational models should provide examples
and/or simulations that show the models’ applicability to a realistic
legal problem or domain. Papers on innovative applications should
describe clearly the motivations behind the project, the techniques
employed, and the current state of both implementation and evaluation.
All papers should make clear their relation to prior work.


Demonstrations

A session will be organized for the demonstration of creative, robust
and practical working applications and tools. Where a demonstration is
not connected to a paper in a track, a two page extended abstract
about the system should be submitted for review by the paper
submission deadline via the conference management system and following
the conference style. Accepted extended abstracts will be published in
the conference proceedings. For those demonstrations that are
connected to a paper in the regular track or innovative applications
track, no separate statement about the demonstration should be
submitted.


ICAIL Workshops and Tutorials

ICAIL 2013 will include workshops and tutorials on the first and last
days. Proposals for workshops and tutorials are invited, and should be
sent to the Program Chair. Tutorials should cover a broad topic of
relevance to the AI and Law community. Proposals should contain enough
information to permit evaluation on the basis of importance, quality,
and community interest. Each workshop should have one or more
designated organizers and a program or organizing committee. Proposals
should be 2 to 4 pages and include at least the following information:

* The workshop or tutorial top

Re: [UAI] This note is neither a conference announcement nor a job posting! It asks a question about learning Bayesian networks.

2012-11-22 Thread Milan Studeny
Dear Richard,
 in recent years, I have devoted a lot of my research effort to the development 
of (theoretical basis for) the 
geometric methods, which can be used to find a global maximum of quality 
criterion (=score) over BN structures. 
This, in the end, seems to lead to the application of methods of (integer) 
linear programming, for which wonderful 
speedy software packages have been developed by specialists in mathematical 
optimization

Thus, I have mentioned that there were some papers devoted to the algorithms
for finding the global maximum and some computational experiments based on that 
have been made. 
These are mainly
- dynamic programming approach by Silander+Myllymakki 
- integer linear programming approach by Jaakkolla, Sontag, Globerson and Meila 
- integer linear programming approach by Cussens (2 papers)
- higly relevant is also the research by C.deCampos +Ji on "pruning"

Of course, my co-authors (in particular S.Lindner) also made some preliminary 
computational experiments.
You may find references to the above mentioned papers (and some description of 
our approach) in a couple of papers:

- M. Studený, J. Vomlel, R. Hemmecke ``A geometric view on learning Bayesian 
network structures''
International Journal of Approximate Reasoning 51 (2010), n.5, 578-586.
a preprint available through staff.utia.cas.cz/studeny/a28.html
[BTW: in this paper a very simple example is given showing that GES can fail to 
find the global maximum]

- R. Hemmecke, S. Lindner, M. Studený ``Characteristic imsets for learning 
Bayesian network structure'' 
   International Journal of Approximate Reasoning 53 (2012) 1336-1349.
a preprint available through staff.utia.cas.cz/studeny/a31.html
[Here you find most of the references]

I hope this helps. Regards from 
 Milan Studeny

PS: in my view, the application of the integer linear programming approach in 
particularly promising
   research direction





Richard E. Neapolitan  napsal:
>   Dear Colleagues,
>   This note concerns learning Bayesian networks from data,  
> an area in which I wrote a book in 2003. However, since then I have not kept 
> that close of a track of  developments in the area. 
> The GES algorithm  assumes the composition property,  
>   and the
> constraint-based PC algorithm andmore 
> advanced constraint-based algorithms  
> assume faithfulness or embedded
> faithfulness.  So none of 
>them would discover a  
> DAG in which  
> two variables 
>  together have
>   an effect on a  
> third variable,   
>but neither of 
>  the variables  has a 
> marginal  effect. I 
> am  wondering if  
> there are any heuristic   
>search 
>  algorithms,  
> in a particular   
>ones implemented   
>in available   
>software, that 
>  address this 
>  situation.  Clearly, 
> there  are 
> modifcations  of 
> these  algorithms 
>  that would do
>   so.
>   Thanks,
>   Rich
>   
>   -- Richard E. Neapolitan, Ph.D., ProfessorDivision of 
> Health and Biomedical Infor