Showing 1–8 of 8 results for author: Georgatos, K

Relevance as Deduction: A Logical View of Information Retrieval

Authors: Gianni Amati, Konstantinos Georgatos

Abstract: The problem of Information Retrieval is, given a set of documents D and a query q, providing an algorithm for retrieving all documents in D relevant to q. However, retrieval should depend and be updated whenever the user is able to provide as an input a preferred set of relevant documents; this process is known as em relevance feedback. Recent work in IR has been paying great attention to models… ▽ More The problem of Information Retrieval is, given a set of documents D and a query q, providing an algorithm for retrieving all documents in D relevant to q. However, retrieval should depend and be updated whenever the user is able to provide as an input a preferred set of relevant documents; this process is known as em relevance feedback. Recent work in IR has been paying great attention to models which employ a logical approach; the advantage being that one can have a simple computable characterization of retrieval on the basis of a pure logical analysis of retrieval. Most of the logical models make use of probabilities or similar belief functions in order to introduce the inductive component whereby uncertainty is treated. Their general paradigm is the following: em find the nature of conditional $d\imp q$ and then define a probability on the top of it. We just reverse this point of view; first use the numerical information, frequencies or probabilities, then define your own logical consequence. More generally, we claim that retrieval is a form of deduction. We introduce a simple but powerful logical framework of relevance feedback, derived from the well founded area of nonmonotonic logic. This description can help us evaluate, describe and compare from a theoretical point of view previous approaches based on conditionals or probabilities. △ Less

Submitted 26 July, 2000; originally announced July 2000.

Comments: 6 pages, Abstract

Report number: TR-1996-29 ACM Class: I.2.3; H.3.0; H.3.3

Journal ref: In F. Crestani and M. Lalmas, editors, Proceedings of the Second Workshop on Information Retrieval, Uncertainty and Logic WIRUL'96, pages 21--26. University of Glasgow, Glasgow, Scotland, 1996

Entrenchment Relations: A Uniform Approach to Nonmonotonicity

Authors: Konstantinos Georgatos

Abstract: We show that Gabbay's nonmonotonic consequence relations can be reduced to a new family of relations, called entrenchment relations. Entrenchment relations provide a direct generalization of epistemic entrenchment and expectation ordering introduced by Gardenfors and Makinson for the study of belief revision and expectation inference, respectively. We show that Gabbay's nonmonotonic consequence relations can be reduced to a new family of relations, called entrenchment relations. Entrenchment relations provide a direct generalization of epistemic entrenchment and expectation ordering introduced by Gardenfors and Makinson for the study of belief revision and expectation inference, respectively. △ Less

Submitted 26 July, 2000; originally announced July 2000.

Comments: 22 pages, a paper in preliminary form that appeared later in conference proceedings

ACM Class: F.4.1; I.2.3

Journal ref: In the Proceedings of the International Joint Conference on Qualitative and Quantitative Practical Reasoning (ESCQARU/FAPR 97), Lecture Notes in Artificial Intelligence, vol. 1244, pages 282-297, Springer-Verlag, 1997

Ordering-based Representations of Rational Inference

Authors: Konstantinos Georgatos

Abstract: Rational inference relations were introduced by Lehmann and Magidor as the ideal systems for drawing conclusions from a conditional base. However, there has been no simple characterization of these relations, other than its original representation by preferential models. In this paper, we shall characterize them with a class of total preorders of formulas by improving and extending Gardenfors an… ▽ More Rational inference relations were introduced by Lehmann and Magidor as the ideal systems for drawing conclusions from a conditional base. However, there has been no simple characterization of these relations, other than its original representation by preferential models. In this paper, we shall characterize them with a class of total preorders of formulas by improving and extending Gardenfors and Makinson's results for expectation inference relations. A second representation is application-oriented and is obtained by considering a class of consequence operators that grade sets of defaults according to our reliance on them. The finitary fragment of this class of consequence operators has been employed by recent default logic formalisms based on maxiconsistency. △ Less

Submitted 26 July, 2000; originally announced July 2000.

Comments: 26 pages, appeared in conference proceedings, contains proofs

ACM Class: F.4.1; I.2.3

Journal ref: In the Proceedings of the European Workshop on Logics in AI (JELIA '96). J.J. Alferes, L.M. Pereira and E. Orlowska, Eds. Lecture Notes in Artificial Intelligence, vol. 1126, pages 176-191, Springer-Verlag, 1996

Modal Logics for Topological Spaces

Authors: Konstantinos Georgatos

Abstract: In this thesis we shall present two logical systems, MP and MP, for the purpose of reasoning about knowledge and effort. These logical systems will be interpreted in a spatial context and therefore, the abstract concepts of knowledge and effort will be defined by concrete mathematical concepts. In this thesis we shall present two logical systems, MP and MP, for the purpose of reasoning about knowledge and effort. These logical systems will be interpreted in a spatial context and therefore, the abstract concepts of knowledge and effort will be defined by concrete mathematical concepts. △ Less

Submitted 26 July, 2000; originally announced July 2000.

Comments: 25 pages, extened abstract of PHD Dissertation

ACM Class: F.4.1

Knowledge Theoretic Properties of Topological Spaces

Authors: Konstantinos Georgatos

Abstract: We study the topological models of a logic of knowledge for topological reasoning, introduced by Larry Moss and Rohit Parikh. Among our results is a solution of a conjecture by the formentioned authors, finite satisfiability property and decidability for the theory of topological models. We study the topological models of a logic of knowledge for topological reasoning, introduced by Larry Moss and Rohit Parikh. Among our results is a solution of a conjecture by the formentioned authors, finite satisfiability property and decidability for the theory of topological models. △ Less

Submitted 26 July, 2000; originally announced July 2000.

Comments: 14 pages

ACM Class: F.4.1

Journal ref: In Knowledge Representation and Uncertainty. M. Masuch and L. Polos, Eds. Lecture Notes in Artificial Intelligence, vol. 808, pages 147-159, Springer-Verlag, 1994

To Preference via Entrenchment

Authors: Konstantinos Georgatos

Abstract: We introduce a simple generalization of Gardenfors and Makinson's epistemic entrenchment called partial entrenchment. We show that preferential inference can be generated as the sceptical counterpart of an inference mechanism defined directly on partial entrenchment. We introduce a simple generalization of Gardenfors and Makinson's epistemic entrenchment called partial entrenchment. We show that preferential inference can be generated as the sceptical counterpart of an inference mechanism defined directly on partial entrenchment. △ Less

Submitted 21 July, 2000; originally announced July 2000.

Comments: 16 pages

ACM Class: F.4.1; I.2.3

Journal ref: Annals Of Pure And Applied Logic, (96)1-3, pages 141-155, 1999

Knowledge on Treelike Spaces

Authors: Konstantinos Georgatos

Abstract: This paper presents a bimodal logic for reasoning about knowledge during knowledge acquisition. One of the modalities represents (effort during) non-deterministic time and the other represents knowledge. The semantics of this logic are tree-like spaces which are a generalization of semantics used for modeling branching time and historical necessity. A finite system of axiom schemes is shown to b… ▽ More This paper presents a bimodal logic for reasoning about knowledge during knowledge acquisition. One of the modalities represents (effort during) non-deterministic time and the other represents knowledge. The semantics of this logic are tree-like spaces which are a generalization of semantics used for modeling branching time and historical necessity. A finite system of axiom schemes is shown to be canonically complete for the formentioned spaces. A characterization of the satisfaction relation implies the small model property and decidability for this system. △ Less

Submitted 21 July, 2000; originally announced July 2000.

Comments: 31 pages

ACM Class: F.4.1; I.2.0

Journal ref: Studia Logica, 1(59), 1997

Relevance Sensitive Non-Monotonic Inference on Belief Sequences

Authors: Samir Chopra, Konstantinos Georgatos, Rohit Parikh

Abstract: We present a method for relevance sensitive non-monotonic inference from belief sequences which incorporates insights pertaining to prioritized inference and relevance sensitive, inconsistency tolerant belief revision. Our model uses a finite, logically open sequence of propositional formulas as a representation for beliefs and defines a notion of inference from maxiconsistent subsets of formul… ▽ More We present a method for relevance sensitive non-monotonic inference from belief sequences which incorporates insights pertaining to prioritized inference and relevance sensitive, inconsistency tolerant belief revision. Our model uses a finite, logically open sequence of propositional formulas as a representation for beliefs and defines a notion of inference from maxiconsistent subsets of formulas guided by two orderings: a temporal sequencing and an ordering based on relevance relations between the conclusion and formulas in the sequence. The relevance relations are ternary (using context as a parameter) as opposed to standard binary axiomatizations. The inference operation thus defined easily handles iterated revision by maintaining a revision history, blocks the derivation of inconsistent answers from a possibly inconsistent sequence and maintains the distinction between explicit and implicit beliefs. In doing so, it provides a finitely presented formalism and a plausible model of reasoning for automated agents. △ Less

Submitted 7 March, 2000; originally announced March 2000.

ACM Class: I.2.3