Uli Sattler

Introducing Defeasibility into OWL Ontologies

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.

Finding Fault: Detecting Issues in a Versioned Ontology

Understanding ontology evolution is becoming an active topic of interest for ontology engineers, e.g., there exist large collaboratively-developed ontologies but, unlike in software engineering, comparatively little is understood about the dynamics of historical changes, especially at a fine level of granularity. Only recently has there been a systematic analysis of changes across ontology versions, but still at a coarse-grained level. The National Cancer Institute (NCI) Thesaurus (NCIt) is a large, collaboratively-developed ontology, used for various Web and research-related purposes, e.g., as a medical research controlled vocabulary. The NCI has published ten years worth of monthly versions of the NCIt as Web Ontology Language (OWL) documents, and has also published reports on the content of, development methodology for, and applications of the NCIt. In this paper, we carry out a fine-grained analysis of asserted axiom dynamics throughout the evolution of the NCIt from 2003 to 2012. From this, we are able to identify axiomatic editing patterns that suggest significant regression editing events in the development history of the NCIt.

Toward cognitive support for OWL justifications

Justifications are the dominant form of explanation for entailments of OWL ontologies, with popular OWL ontology editors, such as Protege 4, providing justification-based explanation facilities. A justification is a minimal subset of an ontology which is sufficient for an entailment to hold; they correspond to the premises of a proof. Unlike proofs, however, justifications do not articulate how their axioms support the entailment. We frequently observe that ontology developers find certain justifications difficult to work with; and while in some cases the sources of difficulty are obvious (such as a large number of axioms), we do not have a good general understanding of what makes justifications easy or difficult for ontology users. In this paper, we present an approach to determining the cognitive complexity of justifications for entailments of OWL ontologies. We describe an exploratory study which forms the basis for a cognitive complexity model that predicts the complexity of OWL justifications, and present the results of validating that model via experiments involving OWL users. This is concluded by an investigation into strategies OWL users apply to support them in understanding justifications. Our contributions include an evaluation of the cognitive complexity model, new insights into the complexity of justifications for entailments of OWL ontologies, a significant corpus with novel analyses of justifications suitable for experimentation, and an experimental protocol suitable for model validation and refinement.

A Multi-reasoner, Justification-Based Approach to Reasoner Correctness

OWL 2 DL is a complex logic with reasoning problems that have a high worst case complexity. Modern reasoners perform mostly very well on naturally occurring ontologies of varying sizes and complexity. This performance is achieved through a suite of complex optimisations (with complex interactions) and elaborate engineering. While the formal basis of the core reasoner procedures are well understood, many optimisations are less so, and most of the engineering details (and their possible effect on reasoner correctness) are unreviewed by anyone but the reasoner developer. Thus, it is unclear how much confidence should be placed in the correctness of implemented reasoners. To date, there is no principled, correctness unit test-like suite for simple language features and, even if there were, it is unclear that passing such a suite would say much about correctness on naturally occurring ontologies. This problem is not merely theoretical: Divergence in behaviour (thus known bugginess of implementations) has been observed in the OWL Reasoner Evaluation (ORE) contests to the point where a simple, majority voting procedure has been put in place to resolve disagreements.

General Terminology Induction in OWL

Automated acquisition, or learning, of ontologies has attracted research attention because it can help ontology engineers build ontologies and give domain experts new insights into their data. However, existing approaches to ontology learning are considerably limited, e.g. focus on learning descriptions for given classes, require intense supervision and human involvement, make assumptions about data, do not fully respect background knowledge. We investigate the problem of general terminology induction, i.e. learning sets of general class inclusions, GCIs, from data and background knowledge. We introduce measures that evaluate logical and statistical quality of a set of GCIs. We present methods to compute these measures and an anytime algorithm that induces sets of GCIs. Our experiments show that we can acquire interesting sets of GCIs and provide insights into the structure of the search space.

Generating Multiple Choice Questions From Ontologies: How Far Can We Go?

Ontology-based Multiple Choice Question (MCQ) generation has a relatively short history. Many attempts have been carried out to develop methods to generate MCQs from ontologies. However, there is still a need to understand the applicability of these methods in real educational settings. In this paper, we present an empirical evaluation of ontology-based MCQ generation. We examine the feasibility of applying ontology-based MCQ generation methods by educators with no prior experience in ontology building. The findings of this study show that this is feasible and can result in generating a reasonable number of educationally useful questions with good predictions about their difficulty levels.

Measuring similarity in ontologies: a new family of measures

Several attempts have been already made to develop similarity measures for ontologies. We noticed that some existing similarity measures are ad-hoc and unprincipled. In addition, there is still a need for similarity measures which are applicable to expressive Description Logics and which are terminological. To address these requirements, we have developed a new family of similarity measures. Two separate empirical studies have been carried out to evaluate the new measures. First, we compare the new measures along with some existing measures against a gold-standard. Second, we examine the practicality of using the new measures over an independently motivated corpus of ontologies.

OWL Reasoning: Subsumption Test Hardness and Modularity

Reasoning with

Practical Defeasible Reasoning for Description Logics

\mathcal {SROIQ(D)}