Kolloquium/Ringvorlesung des Graduiertenkolleg Wissensrepräsentation

Mi, 18.4.2007 17.15-18.45 Prof. Th. Schwentick
Dortmund
"Two variable logics in the presence of an equivalence relation"
Johannisgasse 26, 03-36
Mi, 25.4.2007 15.00-16.30 Prof. Wiebe van der Hoek
Department of Computer Science
University of Liverpool
"Logics for Cooperation, Control and Coalition"
Johannisgasse 26, 03-36
Mi, 9.5.2007 17.15-18.45 Prof. Ulle Endriss
ILLC, University of Amsterdam
"Weighted Propositional Formulas for Preference Representation in Combinatorial Domains"
Johannisgasse 26, 03-36
Mi 23.5.2007 18.15-19.45 Ricardo Baeza-Yates
Web Query Mining
Johannisgasse 26, 03-36
Mi 13.6.2007 17.15-18.45 Steven Krauwer
Universität Utrecht, Niederlande
"Small Languages"
Johannisgasse 26, 03-36
Mi 27.6.2007 x reserviert Johannisgasse 26, 03-36
Mi 4.7.2007 17.15-18.45 Harald Hammarström
Unsupervised Learning of Morphology: Considerations and Applications
Johannisgasse 26, 03-36
Mi 11.7.2007 17.15-18.45 Dr. habil. Carsten Lutz
TU Dresden
"Description Logics for Large-Scale Ontologies"
Johannisgasse 26, 03-36
XX XX Mikel Forcada
Universität Alicante
"Open Source Machine Translation"
Johannisgasse 26, 03-36




   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Prof. Th. Schwentick

Two variable logics in the presence of an equivalence relation



Mortimer showed more than 30 years ago that two-variable logics possesses the Finite Model Property and therefore has a decidable satisfiability problem. Since then, this result has been refined and extended in several ways. In particular, satisfiability on structures with specific properties has been investigated. Of course, the general decidability result can be directly transfered only if the property at hand can be axiomatized with two variables. Recently, the satisfiability problem for two-variable logics on structures with (one or more) equivalence relations has been investigated in more detail. Partially, these investigations were motivated by the fact that adding an equivalence relation to a structure has the same effect as associating a data value to each element and allowing for equality tests on these data values. After a survey of some general results, the focus of the talk will be on recent results for (finite and infinite) strings and trees with one equivalence relation.




Alle Interessenten sind herzlich eingeladen.

zurück
   



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Prof. Wiebe van der Hoek

Logics for Cooperation, Control and Coalition



Joint work with Thomas Agotnes & Mike Wooldridge
We discuss logics for cooperation in which in which agents have control over certain aspects of the world. Moreover, this control can be delegated to other agents. We then discuss Coalition Logic: we show how adding preferences to Pauly's coalition logic CL enables one to express several game theoretical concepts, and we finally show how adding for a restricted form of quantification gives a language that is equally expressive, but exponentially more succingt than CL.




Alle Interessenten sind herzlich eingeladen.

zurück
   



   



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Prof. Ulle Endriss

Weighted Propositional Formulas for Preference Representation in Combinatorial Domains




Many collective decision making problems require individual agents to somehow represent their preferences over alternative agreements. The set of alternatives to be considered often has a combinatorial structure. For instance, in the context of negotiation over indivisible goods, the number of possible allocations of goods to agents will be exponential in the number of goods. Similarly, in the context of voting for a committee (rather than just a single person), the number of alternative committees will be exponential in the number of seats to be filled. This calls for knowledge representation languages that can model typical preference structures in a concise manner. In this talk, I shall discuss a class of logic-based preference languages based on weighted propositional formulas. This is joint work with Yann Chevaleyre, Jerome Lang, and Joel Uckelman.




Alle Interessenten sind herzlich eingeladen.

zurück
     



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Ricardo Baeza-Yates

Web Query Mining




User queries in search engines and Websites give valuable information on the interests of people. In addition, clicks after queries relate those interests to actual content. Even queries without answers imply important missing synonyms or content. In this talk we show several examples on how to use this information to improve the performance of search engines, to recommend better queries, and to improve the information scent of the content of a Website.




Alle Interessenten sind herzlich eingeladen.

zurück
   



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Steven Krauwer

Small Languages



The title of my talk is 'Smaller Languages'. I will ask (but not answer) the question whether there is any hope for the smaller languages, and I will focus on the question how we can try to use technology to create better conditions for the smaller languages. I will discuss some recent and upcoming initiatives such as BLARK, BLARKette and CLARIN and show how they can be used to the benefit of smaller languages.




Alle Interessenten sind herzlich eingeladen.

zurück
   



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Harald Hammarström

Unsupervised Learning of Morphology: Considerations and Applications



Unsupervised learning of morphology is the the topic of my PhD thesis work. I will present a short survey of the quite substantial, but often forgotten, amount of work done in the field since 1955, and highlight what we can learn from it and how to proceed. I will argue that developing a theory of affixation with explicit assumption about frequency patterns is more helpful in getting us closer to the goal. I will present my own embryo to such a theory and show how it can be applied to extract salient affixes. Extraction of affixes is not the same as segmentation, but I will give two applications where this can be used to yield fully unsupervised classifiers (= no thresholds or parametres at all): 1) Decide if two words share the same stem and 2) identify the language of a short specimen (i.e. one word).




Alle Interessenten sind herzlich eingeladen.

zurück
   



   

Kolloquium des Graduiertenkolleg Wissensrepräsentation

Dr. habil. Carsten Lutz

Description Logics for Large-Scale Ontologies



Description logics (DLs) are a family of logic-based knowledge representation formalisms that have recently become a standard as ontology languages. The processing of large ontologies, such as those originating in the life sciences, poses a number of new challenges to DL research. In this talk, I discuss two of these challenges:

(i) to engineer DLs with polytime reasoning problems, and

(ii) to develop a formal treatment of notions such as ontology refinement and modularity.

In particular, Challenge (ii) provides the motivation to study conservative extensions in the context of DLs.




Alle Interessenten sind herzlich eingeladen.

zurück