Knowledge Representation and Reasoning Research Publications

In this paper, we enrich description logics (DLs) with non-monotonic reasoning features in a number of ways. We start by investigating a notion of defeasible conditional in the spirit of KLM-style defeasible consequence. In particular, we consider a natural and intuitive semantics for defeasible subsumption in terms of DL interpretations enriched with a preference relation. We propose and investigate syntactic properties (à la Gentzen) for both preferential and rational conditionals and prove representation results for the description logic ALC. This representation result paves the way for more effective decision procedures for defeasible reasoning in DLs. We then move to non-monotonicity in DLs at the level of entailment. We investigate versions of entailment in the context of both preferential and rational subsumption, relate them to preferential and rational closure, and show that computing them can be reduced to classical ALC entailment. This provides further evidence that our semantic constructions are appropriate in a non-monotonic DL setting. One of the barriers to evaluating performance scalability of rational closure is the abscence of naturally occurring DL-based ontologies with defeasible features. We overcome this barrier by devising an approach to introduce defeasible subsumption into classical real world ontologies. This culminates in a set of semi-natural defeasible ontologies that is used, together with a purely artificial set, to test our rational closure algorithms. We found that performance is scalable on the whole with no major bottlenecks.

@misc{130,
  author = {Katarina Britz and Giovanni Casini and Tommie Meyer and Kody Moodley and U. Sattler and Ivan Varzinczak},
  title = {Rational Defeasible Reasoning for Expressive Description Logics},
  abstract = {In this paper, we enrich description logics (DLs) with non-monotonic reasoning features in a number of ways. We start by investigating a notion of defeasible conditional in the spirit of KLM-style defeasible consequence. In particular, we consider a natural and intuitive semantics for defeasible subsumption in terms of DL interpretations enriched with a preference relation. We propose and investigate syntactic properties (à la Gentzen) for both preferential and rational conditionals and prove representation results for the description logic ALC. This representation result paves the way for more effective decision procedures for defeasible reasoning in DLs. We then move to non-monotonicity in DLs at the level of entailment. We investigate versions of entailment in the context of both preferential and rational subsumption, relate them to preferential and rational closure, and show that computing them can be reduced to classical ALC entailment. This provides further evidence that our semantic constructions are appropriate in a non-monotonic DL setting. One of the barriers to evaluating performance scalability of rational closure is the abscence of naturally occurring DL-based ontologies with defeasible features. We overcome this barrier by devising an approach to introduce defeasible subsumption into classical real world ontologies. This culminates in a set of semi-natural defeasible ontologies that is used, together with a purely artificial set, to test our rational closure algorithms. We found that performance is scalable on the whole with no major bottlenecks.},
  year = {2015},
}

No Abstract

@misc{117,
  author = {Giovanni Casini and Umberto Straccia and Tommie Meyer},
  title = {A Polynomial Time Subsumption Algorithm for EL⊥ under Rational Closure},
  abstract = {No Abstract},
  year = {2015},
}

In recent years, various approaches have been developed for representing and reasoning with exceptions in OWL. The price one pays for such capabilities, in terms of practical performance, is an important factor that is yet to be quantified comprehensively. A major barrier is the lack of naturally occurring ontologies with defeasible features - the ideal candidates for evaluation. Such data is unavailable due to absence of tool support for representing defeasible features. In the past, defeasible reasoning implementations have favoured automated generation of defeasible ontologies. While this suffices as a preliminary approach, we posit that a method somewhere in between these two would yield more meaningful results. In this work, we describe a systematic approach to modify real-world OWL ontologies to include defeasible features, and we apply this to the Manchester OWL Repository to generate defeasible ontologies for evaluating our reasoner DIP (Defeasible-Inference Platform). The results of this evaluation are provided together with some insights into where the performance bottle-necks lie for this kind of reasoning. We found that reasoning was feasible on the whole, with surprisingly few bottle-necks in our evaluation.

@{113,
  author = {Giovanni Casini and Tommie Meyer and Kody Moodley and Ivan Varzinczak and U. Sattler},
  title = {Introducing Defeasibility into OWL Ontologies},
  abstract = {In recent years, various approaches have been developed for representing and reasoning with exceptions in OWL. The price one pays for such capabilities, in terms of practical performance, is an important factor that is yet to be quantified comprehensively. A major barrier is the lack of naturally occurring ontologies with defeasible features - the ideal candidates for evaluation. Such data is unavailable due to absence of tool support for representing defeasible features. In the past, defeasible reasoning implementations have favoured automated generation of defeasible ontologies. While this suffices as a preliminary approach, we posit that a method somewhere in between these two would yield more meaningful results. In this work, we describe a systematic approach to modify real-world OWL ontologies to include defeasible features, and we apply this to the Manchester OWL Repository to generate defeasible ontologies for evaluating our reasoner DIP (Defeasible-Inference Platform). The results of this evaluation are provided together with some insights into where the performance bottle-necks lie for this kind of reasoning. We found that reasoning was feasible on the whole, with surprisingly few bottle-necks in our evaluation.},
  year = {2015},
  journal = {The International Semantic Web Conference},
  month = {11/10-15/10},
}

Representing temporal data in conceptual data models and ontologies is required by various application domains. For it to be useful for modellers to represent the information precisely and reason over it, it is essential to have a language that is expressive enough to capture the required operational semantics of the time-varying information. Temporal modelling languages have little support for temporal attributes, if at all, yet attributes are a standard element in the widely used conceptual modelling languages such as EER and UML. This hiatus prevents one to utilise a complete temporal conceptual data model and keep track of evolving values of data and its interaction with temporal classes. A rich axiomatisation of fully temporised attributes is possible with a minor extension to the already very expressive description logic language DLRUS. We formalise the notion of transition of attributes, and their interaction with transition of classes. The transition specified for attributes are extension, evolution, and arbitrary quantitative extension.

@misc{91,
  author = {Nasubo Ongoma and Maria Keet and Tommie Meyer},
  title = {Transition Constraints for Temporal Attributes},
  abstract = {Representing temporal data in conceptual data models and ontologies
is required by various application domains. For it to be useful for modellers to
represent the information precisely and reason over it, it is essential to have a language
that is expressive enough to capture the required operational semantics of
the time-varying information. Temporal modelling languages have little support
for temporal attributes, if at all, yet attributes are a standard element in the widely
used conceptual modelling languages such as EER and UML. This hiatus prevents
one to utilise a complete temporal conceptual data model and keep track of
evolving values of data and its interaction with temporal classes. A rich axiomatisation
of fully temporised attributes is possible with a minor extension to the
already very expressive description logic language DLRUS. We formalise the
notion of transition of attributes, and their interaction with transition of classes.
The transition specified for attributes are extension, evolution, and arbitrary quantitative
extension.},
  year = {2014},
  journal = {27th International Workshop on Description Logics},
  month = {17/07 - 20/07},
  url = {http://ceur-ws.org/Vol-1193/paper_25.pdf},
}

The introduction of defeasible reasoning in Description Logics has been a main research topic in the field in the last years. Despite the fact that various interesting formalizations of nonmonotonic reasoning for the TBox have been proposed, the application of such a kind of reasoning also to ABoxes is more problematic. In what follows we are going to present the adaptation for the ABox of a classical nonmonotonic form of reasoning, Lehmann and Magidor’s Rational Closure. We present both a procedural and a semantical characterization, and we conclude the paper with a comparison between our and other analogous proposals, and suggesting some possible heuristics useful for the ABox querying procedure.

@misc{114,
  author = {Giovanni Casini and Tommie Meyer and Kody Moodley and Ivan Varzinczak},
  title = {Nonmonotonic reasoning in Description Logics: Rational Closure for the ABox},
  abstract = {The introduction of defeasible reasoning in Description Logics has been a main research topic in the field in the last years. Despite the fact that various interesting formalizations of nonmonotonic reasoning for the TBox have been proposed, the application of such a kind of reasoning also to ABoxes is more problematic. In what follows we are going to present the adaptation for the ABox of a classical nonmonotonic form of reasoning, Lehmann and Magidor’s Rational Closure. We present both a procedural and a semantical characterization, and we conclude the paper with a comparison between our and other analogous proposals, and suggesting some possible heuristics useful for the ABox querying procedure.},
  year = {2013},
}

Conceptual modeling is nowadays mostly done using languages such as Entity-Relationship (ER) Models, Unified Modeling Language (UML), and Object-Role Modeling (ORM). These models are used to depict the ontological organization of relevant concepts or entities. Such models share a common modeling approach, based on the notions of class or entity and the relations or associations between classes or entities. Recent developments in knowledge representation using logic-based ontologies have created new possibilities for conceptual data modeling. It also raises the question of how existing conceptual models using ER, UML or ORM could be translated into Description Logics (DLs), a family of logics that have proved to be particularly appropriate for formalizing ontologies and reasoning about them. Given a conceptual data model, two assumptions are usually made that are not explicitly stated but need to be clarified for its DL translation: (1) disjointness assumption: all the classes are to be assumed pairwise disjoint if not specified otherwise; and (2) covering assumption: the content of every class must correspond to the union of its immediate subclasses (this includes the assumption that we do not consider anything apart from what is expressed in the model). In this paper we propose two simple procedures to assist modelers with integrating these assumptions into their models, thereby allowing for a more complete translation into DLs.

@{7,
  author = {Giovanni Casini and Aurona Gerber and Tommie Meyer},
  title = {A Note on the Translation of Conceptual Data Models into Description Logics: Disjointness and Covering Assumptions},
  abstract = {Conceptual modeling is nowadays mostly done using languages such as Entity-Relationship (ER) Models, Unified Modeling Language (UML), and Object-Role Modeling (ORM). These models are used to depict the ontological organization of relevant concepts or entities. Such models share a common modeling approach, based on the notions of class or entity and the relations or associations between classes or entities. Recent developments in knowledge representation using logic-based ontologies have created new possibilities for conceptual data modeling. It also raises the question of how existing conceptual models using ER, UML or ORM could be translated into Description Logics (DLs), a family of logics that have proved to be particularly appropriate for formalizing ontologies and reasoning about them. Given a conceptual data model, two assumptions are usually made that are not explicitly stated but need to be clarified for its DL translation:
(1) disjointness assumption: all the classes are to be assumed pairwise disjoint if not specified otherwise; and (2) covering assumption: the content of every class must correspond to the union of its immediate subclasses (this includes the assumption that we do not consider anything apart from what is expressed in the model). In this paper we propose two simple procedures to assist modelers with integrating these assumptions into their models, thereby allowing for a more complete translation into DLs.},
  year = {2012},
  journal = {South African Institute for Computer Scientists and Information Technologists (SAICSIT)},
}

In this paper we describe a research project to develop an optimal information retrieval system in an Information Operations domain. Information Operations is the application and management of information to gain an advantage over an opponent and to defend one’s own interests. Corporations, governments, and military forces are facing increasing exposure to strategic information-based actions. Most national defence and security organisations regard Information Operations as both a defensive and offensive tool, and some commercial institutions are also starting to recognise the value of Information Operations. An optimal information retrieval system should have the capability to extract relevant and reasonably complete information from different electronic data sources which should decrease information overload. Information should be classified in a way such that it can be searched and extracted effectively. The authors of this paper have completed an initial phase in the investigation and design of a knowledge system that can be used to extract relevant and complete knowledge for the planning and execution of Information Operations. During this initial phase of the project, we performed a needs analysis and problem analysis and our main finding is the recommendation of the use of logic-based ontologies: it has the advantage of an unambiguous semantics, facilitates intelligent search, provides an optimal trade-off between expressivity and complexity, and yields optimal recall of information. The risk of adopting this technology is its status as an emerging technology and therefore we include recommendations for the development of a prototype system.

@{465,
  author = {Louise Leenen and Ronell Alberts and Katarina Britz and Aurona Gerber and Tommie Meyer},
  title = {Developing a Knowledge System for Information Operations},
  abstract = {In this paper we describe a research project to develop an optimal information retrieval system in an Information Operations domain. Information Operations is the application and management of information to gain an advantage over an opponent and to defend one’s own interests. Corporations, governments, and military forces are facing increasing exposure to strategic information-based actions. Most national defence and security organisations regard Information Operations as both a defensive and offensive tool, and some commercial institutions are also starting to recognise the value of Information Operations. An optimal information retrieval system should have the capability to extract relevant and reasonably complete information from different electronic data sources which should decrease information overload. Information should be classified in a way such that it can be searched and extracted effectively. The authors of this paper have completed an initial phase in the investigation and design of a knowledge system that can be used to extract relevant and complete knowledge for the planning and execution of Information Operations. During this initial phase of the project, we performed a needs analysis and problem analysis and our main finding is the recommendation of the use of logic-based ontologies: it has the advantage of an unambiguous semantics, facilitates intelligent search, provides an optimal trade-off between expressivity and complexity, and yields optimal recall of information. The risk of adopting this technology is its status as an emerging technology and therefore we include recommendations for the development of a prototype system.},
  year = {2011},
  journal = {6th International Conference on Information Warfare and Security},
  month = {17/03-18/03},
  address = {Washington D.C., United States of America},
  isbn = {978-1-906638-93-1},
  doi = {10204/4951},
}

n.a.

@misc{462,
  author = {Johann Bergh and Aurona Gerber and Tommie Meyer and Lynette van Zijl},
  title = {Path analysis for ontology comprehension},
  abstract = {n.a.},
  year = {2011},
  journal = {Seventh Australasian Ontology Workshop},
  month = {12/11},
}