DIMACS-LAMSADE partnership
 

Two leading research centers, DIMACS (the Center for Discrete Mathematics and Theoretical Computer Science, based at Rutgers University), and LAMSADE (the Laboratoire d'Analyse et Modélisation de Systèmes pour l'Aide à la Décision, based at Université Paris IX - Dauphine) have initiated an international collaboration organized around modern computer science applications of methods developed by decision theorists, in particular methods involving consensus and associated order relations. The project seeks to explore the connections between computer science and decision theory, develop new decision-theory-based methodologies relevant to the scope of modern CS problems, and investigate their applications to problems of computer science and also to problems of the social sciences which could benefit from new ideas and techniques. The project features exchange visits of graduate students and junior researchers and several workshops aimed at introducing a broader community to the topics of the collaboration. After the first three years and two successful workshops (put link here to 2004 and 2006) the project is now more specifically oriented to Algorithmic Decision Theory which is also the main subject of the COST Action IC0602 to which both laboratories participate.

Intellectual Merit : The project emphasizes computer science problems that arise in meta-search (combining the outputs of several search engines); collaborative filtering (using learning algorithms to make recommendations for books, movies, etc.); finding centrally located items in large databases, in particular biological ones and ones that arise in homeland security applications; combining ratings of software or hardware; in distributed computing when some processors fail and we seek to reach ``agreement" among the remaining processors; finding efficient ways to compute consensus functions; and learning about individuals' preferences as they are revealed gradually in situations of economic cooperation and competition using the Internet.

Broader Impacts: The project is expected to have impact well beyond the small number of people participating in the exchange visits through a center-to-center exchange that will involve a large number of scientists associated with the two centers in the related scientific discussions and interchanges. Moreover, the impact will be broadened through the workshops that will introduce many of those both associated with and outside the two collaborating centers to this new field. The results should be broadly useful in emerging information technology applications, in areas of economics and political science where methods of decision theory have traditionally been applied, and in new areas of application of decision theory such as to problems of epidemiology and bio-terrorism.

If you want more details on the original project see here [.pdf]

For more details on Algorithmic Decision Theory look here.