The approach of performing research and collecting data for analysis can be qualitative or quantitative. The choice of approach depends upon the problem at hand. A quantitative research approach is associated with the positivist research tradition and applies scientific methods to social science. This research is concerned with the collection of quantitative data in the form of numbers, collected by techniques such as questionnaires and other measurement instruments. This approach considers that knowledge can be based on what can be objectively observed and experienced . On the other hand qualitative research is concerned with developing explanations of social phenomena and is concerned with data in the form of words, collected by techniques such as interviews and observation. The qualitative research paradigm is associated with the interpretivist research tradition and concerned with the opinions, experiences and feelings of individuals producing subjective data ,.
There are some research questions that cannot easily be answered by a quantitative research design but can be answered by qualitative research methods instead. Qualitative research allows the researcher to get much richer answers to questions and give valuable insights which might have been missed by other methods. It not only provides valuable information on its own but can be used to complement quantitative research methods .
The value of data depends on its trustworthiness that is validity and reliability of the data. The trustworthiness is high if quantitative and qualitative approaches of data collection and analysis can be combined rather than being used separately .
In our research the focus is on the collection of both qualitative and quantitative data, so we will use both kinds of research methods.
Outline of Thesis Work
The work starts with the literature review. Our first task was to study the current patterns that exist in other fields and compare them to ontology design patterns’ templates to analyze the difference. In the next phase, we conducted two different types of online surveys.
The aim of the first survey was to determine how well the current ODP template supports the understanding and usage of Content ODPs. The participants of the online survey were the students who attended the Information Logistics course at Jönköping University, Sweden. The reason for their selection was because they were familiar with ODPs and used them for ontology development during the course. The questionnaire was sent to forty students but we got response from seventeen of them.
In the survey, a description was given of one of the ontology design patterns, called AgentRole (see Appendix 8.1.1). The pattern and its description were copied from an online portal, i.e. the ODP portal at http://ontologydesignpatterns.org. The participants were required to read the description and understand different parts of the pattern and then answer a questionnaire based on their observations. The questionnaire of the survey was designed to collect wide range of opinion in the form of both open-ended and closed-ended questions. The first four questions of the survey were closed-ended which were designed to specifically cover each part of template while the purpose of next three questions was to allow participants to provide suggestions for the improvement of current template.
The aim of second survey was to get experts opinions on the current structure of the Ontology design pattern template. The participants of the online survey were the members of the ODP portal quality committee. In total, nine members participated in the survey. The survey consisted of two parts (see Appendix 8.1.2):
In Part 1, the participants were asked to select a set of patterns for an ontology engineering problem, based on seeing only a part of the complete template for content ODPs. The results from the first survey showed that the Graphical Representation, General Description and Scenario were the three most important parts in the current template of ODP. The purpose of this section was to identify the most important part among them.
The questionnaire in Part 2 was the same as in the first survey where participants were asked to give their opinion with respect to the content and structure of the current template. The reason for conducting the same survey was to observe the difference in opinion between novice and expert users in understanding ODPs.
The techniques used for data collection are literature review and survey. The literature review stage can be considered as a qualitative technique as it develops a conceptual framework before collection of further data . Surveys were set up after a period of examining the relevant literature, i.e. conducting a literature review.
Survey research involves the collection of primary data from all or part of a population. There are different types of surveys, each encompasses a variety of data collection techniques e.g. questionnaires, interviews and observation. We have used web surveys to conduct our research and questionnaire as a data collection technique. A questionnaire is a combination of close ended or open ended questions or both. A closed ended question requires the respondent to select an answer from a number of options while an open ended question asks the respondent to devise his own answer . The questionnaires used in both surveys were a combination of open ended and closed ended questions.
The responses from the closed ended questions were measured by ranking the answers. To cover whole range of responses, the importance was measured by categorizing the answers in the form of Most Important, Important, Neither Important nor Unimportant, Not Important and Not Important at all.
In the Part 1 of the second survey, the participants were to select a correct set of patterns for the four competency questions based on seeing only one part of the pattern. The survey randomly generated a single part (i-e Graphical Representation, General Description or Scenario) for each participant. There were total nine participants and each part of the pattern received three answers.
Evaluation methods are used to validate the research results. In our study, evaluation was done on the results of the comparison of different patterns and the survey results. Templates of the patterns were compared to identify the difference and similarities in their presentation. Each part of the templates was studied with respect to its objective and the content provided in that part.
Evaluation of the survey results was done by analyzing the responses of both open ended and closed ended questions. The closed ended questions were analyzed by calculating the frequency of responses for each question to identify the most important part. For open ended questions, all the comments from the participants were read through and categories for responses were developed. After studying the data, trends were identified and the results were summarized to represent general opinion. Finally, we used the notions of external and internal validity to analyze the results of the study.
1.2 PURPOSE AND OBJECTIVES
1.3 RESEARCH QUESTIONS
1.6 THESIS OUTLINE
2 Theoretical Background
2.1.1 Relation to Other Models
2.1.2 Main Concepts of Ontologies
2.1.3 Ontology Levels
2.1.4 Ontology Language
2.1.5 Ontology Engineering
2.2 KNOWLEDGE REUSE
2.2.1 Ontology Reuse
2.4 ONTOLOGY PATTERNS
2.4.1 Classification of Ontology Pattern
2.5 ONTOLOGY DESIGN PATTERNS
2.6 TEMPLATE OF CONTENT ONTOLOGY DESIGN PATTERNS
3 Research Methodology
3.1 RESEARCH METHOD
3.2 OUTLINE OF THESIS WORK
3.2.1 First Survey
3.2.2 Second Survey
3.3 DATA COLLECTION
3.4 EVALUATION METHODS
4.1 FIRST SURVEY
4.2 SECOND SURVEY
4.2.1 Part 1
4.2.2 Part 2
5 Analysis and Discussion
5.1 COMPARISON OF PATTERN TEMPLATES
5.2 SURVEY ANALYSIS
5.2.1 First Survey
5.2.2 Second Survey
GET THE COMPLETE PROJECT