In this work we focus on situation calculus action theories over generalized databases with equality constraints, here called GFDBs, which are able to finitely represent complete information over a possibly infinite number of objects. We contribute with the following: i) we show that GFDBs characterize the class of definitional KBs and that they are closed under progression; ii) we show that temporal projection queries are decidable for theories with an initial KB expressed as a GFDB, which we call GFDB-BATs; iii) we extend the notion of boundedness to allow for infinite objects in the extensions of fluents and prove that a wide class of generalized projection queries is decidable for GFDB-BAT under a restriction we call C-boundedness; iv) we show that checking whether C-boundedness holds for a given bound is decidable. The proposed action theories are to date the most expressive ones for which there are decidable methods
for computing both progression and generalized projection.
%0 Conference Paper
%1 patrizi2014action
%A Patrizi, Fabio
%A Vassos, Stavros
%B Logics in Artificial Intelligence - 14th European Conference, JELIA
%D 2014
%K Genaralized-Databases Progression Projection Situation-Calculus optique-project
%P 472-485
%R 10.1007/978-3-319-11558-0_33
%T Action Theories over Generalized Databases with Equality Constraints
%X In this work we focus on situation calculus action theories over generalized databases with equality constraints, here called GFDBs, which are able to finitely represent complete information over a possibly infinite number of objects. We contribute with the following: i) we show that GFDBs characterize the class of definitional KBs and that they are closed under progression; ii) we show that temporal projection queries are decidable for theories with an initial KB expressed as a GFDB, which we call GFDB-BATs; iii) we extend the notion of boundedness to allow for infinite objects in the extensions of fluents and prove that a wide class of generalized projection queries is decidable for GFDB-BAT under a restriction we call C-boundedness; iv) we show that checking whether C-boundedness holds for a given bound is decidable. The proposed action theories are to date the most expressive ones for which there are decidable methods
for computing both progression and generalized projection.
@inproceedings{patrizi2014action,
abstract = {In this work we focus on situation calculus action theories over generalized databases with equality constraints, here called GFDBs, which are able to finitely represent complete information over a possibly infinite number of objects. We contribute with the following: i) we show that GFDBs characterize the class of definitional KBs and that they are closed under progression; ii) we show that temporal projection queries are decidable for theories with an initial KB expressed as a GFDB, which we call GFDB-BATs; iii) we extend the notion of boundedness to allow for infinite objects in the extensions of fluents and prove that a wide class of generalized projection queries is decidable for GFDB-BAT under a restriction we call C-boundedness; iv) we show that checking whether C-boundedness holds for a given bound is decidable. The proposed action theories are to date the most expressive ones for which there are decidable methods
for computing both progression and generalized projection.},
added-at = {2014-10-15T16:15:41.000+0200},
audience = {academic},
author = {Patrizi, Fabio and Vassos, Stavros},
biburl = {https://www.bibsonomy.org/bibtex/273189371bb8894e34d2c03897bc1d696/savo.fabio},
booktitle = {Logics in Artificial Intelligence - 14th European Conference, JELIA},
doi = {10.1007/978-3-319-11558-0_33},
interhash = {9b490f336c8fcff199c8e514f1e35206},
intrahash = {73189371bb8894e34d2c03897bc1d696},
keywords = {Genaralized-Databases Progression Projection Situation-Calculus optique-project},
openaccess = {No},
pages = {472-485},
partneroptique = {UNIROMA1},
timestamp = {2016-12-01T18:57:14.000+0100},
title = {Action Theories over Generalized Databases with Equality Constraints},
wpoptique = {WP4},
year = 2014,
yearoptique = {Y2}
}