School of Computing FACULTY OF ENGINEERING Developing a methodology for building small scale domain ontologies: HISO case study Ilaria Corda PhD student School of Computing University of Leeds
Dec 30, 2015
School of ComputingFACULTY OF ENGINEERING
Developing a methodology for building small scale domain ontologies:
HISO case study
Ilaria Corda
PhD student
School of Computing
University of Leeds
Research Problem and QuestionsTerminology issues: IEEE definitionsExisting Methodologies and MethodsHISO methodologySummary
Research Problem:
•Current methods and methodologies are not appropriate for small scale domain ontologies
Research questions:
•Which methodological principles can be followed for developing a History of Science ontology (HISO)?
•How should a methodology be scaled to be suitable for a small domain ontology?
•What degree of involvement should the domain expert have?
Our goal is:
To re-use and deploy relevant elements and aspects of existing methods and methodologies for building a small scale domain
ontology
Some terminology [IEEE,1990; IEEE 1995]
Methodology:
“comprehensive, integrated series of techniques creating a general systems theory of how class of
thought-intensive work ought be performed”
Methods:
“orderly process or procedure used in engineering of a product or performing a service”
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and MethodsHISO methodologySummary
METHONTOLOGY (based on the IEEE 1074-1995)-Chemical & Crystalline Elements, Environmental pollutants ontology
Gruninger & Fox’s Methodology -TOVE Toronto Virtual Enterprise project(Enterprise Design methodology, Project ontology, scheduling ontology)
Uschold & King’s Method -Enterprise modelling process ontology
Ordnance Survey’s Methodology (Kanga)-Hydrology ontology (active involvement of domain experts)
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
METHONTOLOGY [Fernandez-Lopez et al.; 99]
Ontology specification activity:
-Purpose
-Scope
-Granularity
-Sources used (e.g. interviews with experts)
Knowledge acquisition:
-Meetings with experts looking at coarse grained knowledge
-In-depth study of the documentation
-Top down knowledge elicitation (from general to specific)
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Conceptualization: it converts informal views of the domain into semi-formal specifications (aka intermediate representations)
-Building glossary of terms (e.g. concepts, instances)
-Constructing concept classification trees
-Defining binary relations diagrams
-Describing axioms & rules
Evaluation & Implementation
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Gruninger and Fox [Gruninger &Fox, 95]
•Identifying a Motivating scenario: to define the scope
•Defining (Informal) Competency Questions: NL questions expressing the requirements (throughout the whole ontology development)
•Expressing Competency Questions & Specifying Axioms in FOL
•Achieving formulation of completeness for CQ
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Ordnance Survey methodology(Kanga) [Kovacs et. al; 06]
•Identifying scope, purpose & requirement of the ontology;
•Gathering source knowledge and relevant documents (books, articles, dictionaries, web resources);
•Building and populating a glossary of terms;
•Converting knowledge contained in the glossary into structured English sentences;
•Evaluation and documentation.
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Uschold and King [Uschold & King; 95]
•Identifying the scope
•Identifying a range of intended users (audience)
•Building the ontology
•Identifying the acquiring the domain knowledge (capture)
•Expressing concepts & relations in a formal language (coding)
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Coding consists of 3 main tasks: •committing to the vocabulary;•choosing a formal language to express concepts and their relations;• programming.
•Reusing existing ontologies (integrating)
•Evaluating the ontology
•Documenting the ontology (for knowledge sharing)
Research Problem and Questions
Terminology issues: IEEE definitions
Existing Methodologies and MethodsHISO methodologySummary
Stages of ontology construction (HISO): Pre-conceptualization
• defining scope and purpose
Conceptualization
• identifying concepts and relations,
• adding time and event descriptions
Logical representation and coding
• adding axioms,
• implementing inference rules for query answering
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and Methods
HISO methodologySummary
Pre-conceptualization (mostly domain expert):
•Providing a high level description of the domain and its features;
•Building a scenario and identifying intended users (Uschold & King)
•Identifying a range of potential informal competency questions to be addressed.
In parallel we start collecting references for:
•Eliciting the main characteristics of the domain
•Investigating existing projects involving the use of ontologies in historical domains.
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and Methods
HISO methodologySummary
Conceptualization:•Identifying main concepts and relations (inspired by METHONTOLOGY & KANGA)
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and Methods
HISO methodologySummary
Logical representation and coding
•Choosing the ontology language
•Converting glossary knowledge into formal representation
•Querying and reasoning about the domain
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and Methods
HISO methodologySummary
Problem
Existing Ontology methods and Methodologies are mostly suitable for large scale ontology development
Outcomes
Domain ontology methodology (History of Science) as a result of relevant elements and aspects taken from existing methods/methodologies
Potentially, the outlined domain ontology methodology can be domain independent
Research Problem and Questions
Terminology issues: IEEE definitionsExisting Methodologies and MethodsHISO methodology
Summary