NSF / ITR – EC / IST 1 United States National Science Foundation and European Commission

نویسندگان

  • Judith S. Olson
  • Dan Cooney
  • Gary M. Olson
  • Erik Hofer
  • Nathan Bos
  • Jude Yew
  • Abigail Potter
چکیده

This report describes a collaboration between researchers at the Technion in Haifa and the University of Southern California’s Information Sciences Institute in Los Angeles. The goal of the collaboration is to jointly construct an email analysis and routing system that will help city governments more rapidly deal with communications from their citizens. This system, or some variant of it, is planned to be incorporated into the EU-funded research project QUALEG, to which the Haifa partners belong; the US partners are funded independently for this collaboration by the NSF. Since the project is only halfway along, this report describes technical results to date and discusses administrative and legal issues that affect the collaboration. 1 Part I: Technical Collaboration 1.1 Partners and Overall Project Structure This collaboration exists between researchers of the Natural Language Group at the University of Southern California’s Information Sciences Institute (USC/ISI) in Los Angeles, USA, and the Faculty of Industrial Engineering and Management at the Technion in Haifa, Israel. The NL Group at USC/ISI (http:/www.isi.edu/natural-language) has a long history of research in most aspects of human language technology, including information extraction, machine translation, text summarization, question answering, ontology construction, parsing, generation, and other aspects. Contact: Dr. Eduard Hovy, [email protected]; +1-310-448-8731. The Technion researchers are part of the EU-funded QUALEG Project (not a Network of Excellence), which falls in the EU’s eGovernment Programme (IST Strategic Objective 2.3.1.9 “Networked businesses and governments”; see http://www.cordis.lu/ist/so/business-govt/). QUALEG comprises 11 partners (4 research, 4 commercial, and 3 city government): • Israel (research): Technion — Israel Institute of Technology (http://www.technion.ac.il/). Contact: Dr. Avigdor Gal, [email protected]; +972-4-8294425 • Italy (research): Associazione Impresa Politecnico (http://www.impresapolitecnico.polimi.it/) • France (research): Ecole des Hautes Etudes Commerciales (http://www.hec.fr/) • Poland (research and conference organizers): Stowarzyszenie “Miasta w Internecie” (http://www.mwi.pl/) • France (commercial): Business Flow Consulting (http://www.business-flow.com/) contact: Norbert Benamou, tel +33 1 41879620 — Project Lead • France (commercial): SQLI (http://www.sqli.com/)

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Optimal Rates for Regularization Operators in Learning Theory

We develop some new error bounds for learning algorithms induced by regularization methods in the regression setting. The “hardness” of the problem is characterized in terms of the parameters r and s, the first related to the “complexity” of the target function, the second connected to the effective dimension of the marginal probability measure over the input space. We show, extending previous ...

متن کامل

Adaptation for Regularization Operators in Learning Theory

We consider learning algorithms induced by regularization methods in the regression setting. We show that previously obtained error bounds for these algorithms using a-priori choices of the regularization parameter, can be attained using a suitable a-posteriori choice based on validation. In particular, these results prove adaptation of the rate of convergence of the estimators to the minimax r...

متن کامل

Computer Science and Artificial Intelligence Laboratory Some Properties of Empirical Risk Minimization Over Donsker Classes

We study properties of algorithms which minimize (or almost-minimize) empirical error over a Donsker class of functions. We show that the L2-diameter of the set of almost-minimizers is converging to zero in probability. Therefore, as the number of samples grows, it is becoming unlikely that adding a point (or a number of points) to the training set will result in a large jump (in L2 distance) t...

متن کامل

Computer Science and Artificial Intelligence Laboratory Risk Bounds for Regularized Least-squares Algorithm with Operator-valued Kernels

We show that recent results in [3] on risk bounds for regularized least-squares on reproducing kernel Hilbert spaces can be straightforwardly extended to the vectorvalued regression setting. We first briefly introduce central concepts on operatorvalued kernels, then we show how risk bounds can be expressed in terms of a generalization of effective dimension. This report describes research done ...

متن کامل

Computer Science and Artificial Intelligence Laboratory Empirical Effective Dimension and Optimal Rates for Regularized Least Squares Algorithm

This paper presents an approach to model selection for regularized least-squares on reproducing kernel Hilbert spaces in the semi-supervised setting. The role of effective dimension was recently shown to be crucial in the definition of a rule for the choice of the regularization parameter, attaining asymptotic optimal performances in a minimax sense. The main goal of the present paper is showin...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2005