XML Labelling Scheme Methodology - Research Paper Example

XML LABELLING SCHEME METHODOLOGY 1 Research Process The research’s methodology was directly influenced by the research process onion developed by Saunders et al (2003). This research process comprises five major modalities that influence the overall methodology to the research. These five major modalities are research philosophy, research approach, research strategy, time horizons, and data collection methods. The methods followed in the usage of the research process onion have been summarised in the diagram below and subsequently explained in detail. Adopted from Saunders et al. (2003) 1.2 Research Philosophy The research philosophy basically explains how available knowledge influences the research and how the research seeks to develop new line of knowledge (Diriwächter & Valsiner, 2006). Saunders et al. (2003) therefore posited that the research philosophy shows the researcher’s overall perception of the way knowledge is constructed. To use and construct knowledge for any research of this nature, three major types of research philosophies may be selected from. These are positivism, realism and interpretivism (Remenyi, Williams, Money & Swartz, 2008). The current research made use of the positivism research philosophy. The positivism research philosophy has been explained as an approach to knowledge where the researcher uses scientific reasoning and law-like generalisations in the knowledge construction process (Adams & Schvaneveldt, 2011). This means that using positivism required the researcher to base on scientific processes that are generally empirical and evidence based in showing how the GroupBased scheme could improve on XML labelling (Alstrup & Rauhe, 2002). It was also important to make use of law-like generalisations which could be interpreted in the form of hypotheses that could easily be tested for its validity. In keeping with the research motivation explained above, the hypothesis that this research seeks to assess is: Applying a second layer of labels and grouping the nodes based on the parent-child relationship may facilitate node insertions in dynamic XML data in an efficient way, offering inexpensive labels with an adequate label size growth rate in which it is easy to maintain structural relationships, as well as improved query performance. The rationale that influenced the selection and use of positivism was the need to ensuring that the findings that resulted from the study could easily be assessed by other for validity and authenticity. This is because Green, Johnson and Adams (2006) saw the positivism research philosophy as a very transparent structure that enhances the replication of findings from the study. Because of the scientific reasoning and law-like generalisations, it is always possible that the researcher’s reasons for drawing conclusions on the hypothesis can be easily tested. 1.3 Research Approach The research approach generally describes the means by which the researcher will go about the implementation of the research philosophy. In the current context, the research approach was set to know the way by which the researcher would test the hypothesis set as part of the positivism research philosophy. In research practice, two major research approaches are known, which are deductive and inductive approaches. Given (2008) explained that the deductive research approach is highly suitable for scientific research as it ensures that the researcher develops a hypothesis and systematically tests the hypothesis in establishing a theory. This means that for deductive approach to be used effectively, the researcher must approach data collection from the known to the unknown. This is because the hypothesis is supposedly set based on what the researcher already knows from preliminary data collection exercises (Hart, 2008). Based on the hypothesis, the researcher then establishes a theory that is not known or is relatively new in the field of study. Given the fact that the research used the positivism research philosophy which makes the study a scientific research, the deductive approach became the main approach that underlined the performance of the research. Throughout this research, the above hypothesis is tested based on the deductive research approach which originates from a specific case and proceeds to derive generalizations and theories (Jebreen, 2012, Meheus and Nickles, 2009). In deductive research, the propositions made in the beginning in the study only support the truth of the conclusion but do not guarantee it. As such, a deductive researcher conducts the study cognizant that the conclusion might not be true. A deductive proposition helps the researcher to derive universal theories or statements. Strong deductive propositions increase the probability of the conclusion being true but they do not confirm that truth (Khan and Ullah, 2010). Deductive reasoning is chosen because it takes into account the impact of researcher bias on the outcome of the study (Sans and Laurent, 2008). This is important because different researchers will have different notions and orientations on a given subject, such as in various labelling schemes that cope with dynamic XML documents from different perspectives, and this affects the outcome of studying such a subject (HAMMAWA and SAMPSON, 2011, Stadler, 2004). The fact that deductive research helps in developing a solution to a specific problem is another reason why it is an appropriate approach to use in this research (Khan and Ullah, 2010, Lorenz et al., 2013); as it can also be used to generate recommendations on how to improve certain techniques. Inductive research was however not excluded entirely from the research. Saunders, Lewis and Thornhill (2003) explained an inductive research approach as one in which the researcher formulates the research theory through the critical evaluation of available research variables. This means that instead of using hypothesis as the main route to forming conclusions and theories, the researcher in an inductive study modifies various research variables based on accumulated findings (Cooper, 2008). Aspects of inductive approach were used in the study even though they did not form the main basis on which conclusions were drawn for the study. for example knowledge of what has already been studied on XML labelling schemes in literature were used to set themes or research variables based on which literature was reviewed in the study. The review of literature however did not become the main basis for drawing conclusions but served as a guide for discussing the researcher’s own findings gathered through the deductive approach. In effect, both inductive and deductive approaches were used but with major emphasis and focus on deductive approach so as to maintain the scientific nature of the study. 1.4 Research Strategy The research strategy gives the underling approach used by the researcher in collecting data (Hunter & Leahey, 2008). In this, as many as six possible research strategies are recommended by Saunders et al (2003). Of these, experiment was selected as the most appropriate for this research. The major rationale for selecting experiment is due to its direct relations with the positivism research philosophy. It would be noted that the positivism research philosophy renders the study to be highly scientific in nature (Robson, 2012). Meanwhile, when used as a research strategy, an experiment requires the researcher to engage in the systematic manipulation of controlled testing with the aim of understanding a casual process (Kasim, Alexander & Hudson, 2010). The aim with which experiment is used as research strategy is to ensure that the researcher can manipulate variables and controls with the aim of measuring any changes that may occur in the variables (Moballeghi & Moghaddam, 2008). In the context of the current study, the researcher was concerned with knowing the changes that would occur to GroupBased scheme as a means of improving XML labelling by providing a scheme that deals with insertions without having to re-label and sacrificing the queries’ performance, construction time and memory usage. This means that the GroupBased scheme was the independent variable based on which dependent variables including query performance, construction time and memory usage were all tested. 1.5 Time Horizon The time horizon basically shows the duration or period within which the phenomena or variables of the study are experimented on (Sapsford & Jupp, 2012). In literature, two major time horizons were identified which are longitudinal time horizon and cross-sectional time horizon. Whereas longitudinal time horizon examines a situation or phenomenon over a given period of time, cross-sectional time horizon focuses on a particular moment (Dellinger & Leech, 2007). Cross-sectional time horizon was used or this study as design and implementation specification was developed purposely for the research. These design and implementation specifications ensured that the performance of GroupBased scheme will be tested over a very specific time frame to know the impact of re-labelling on queries’ performance, construction time and memory usage. 1.6 Data Collection Method The overall data collection method used was observations. Observation has been described as a systematic collection of data from a research setting or an experiment through visual interpretation of findings (Creswell, 2007). This explanation given to observation directly explains the rationale for selecting observation as it relates very directly with the experiment research strategy used. To use observation as part of the positivism research philosophy where a hypothesis was developed and gradually tested, a number of processes were followed. These processes have been itemised below. 1.6.1 Formulating a Tentative Hypothesis Based on the observations and the identified patterns, a tentative hypothesis is formulated. In this thesis, after analysing the existing XML labelling schemes (see Chapter 3) and providing explanations for the detected patterns and problems, more general theory is formulated and the research hypothesis is presented (see Section 1.3). As a result, a new labelling scheme called ‘Groupbased’ is proposed (see Chapter 4) that may have the potential to improve the performance of current XML labelling schemes. 1.6.1 Observation and Patterns Identification The first phase of the deductive approach begins by collecting data that is related to the research area and observes them to highlight what patterns or meaning can be extracted from them in order to identify the problem under study. In this thesis, the research problems emerged from a critical investigation of the existing XML labelling schemes (see Chapter 3), which results in determining the research aims and objectives. 1.6.2 Testing the Hypothesis In the third phase of the deductive approach the hypothesis would be subjected to tests to see whether it is verifiable. This thesis relies on the testing and assessment of an empirical implementation to explore the research hypothesis (see Chapters 5 and 6). This process comprises a number of aspects, including appraisal of the original labelling time and size to determine the extent to which the suggested scheme can be applied in both static and dynamic XML documents; measurement of the amount of time necessary for the identification of structural relationships prior to and following insertions; the impact of various types of insertion on the scheme with regard to the size of the labels and time measurements; and the response time for queries prior to and following insertions (see Chapter 7). 1.6.3 Develop a Theory Evaluating the experiments that used to test the hypothesis and their results (see Chapter 8) will help in generalising a theory and determining the main contributions and limitations of this research (see Chapter 9). As obtaining persistent results after several tests would mean that the hypothesis is supported. Inconsistent results would mean that the hypothesis needs to be changed or rejected. However, in the end a general theory or statement should be defined that can help explain similar cases (Li et al., 2014, Weinstein, 2010). References Adams, G. & Schvaneveldt, J. (2011). Understanding Research Methods. New York: Longman. Alstrup, S. & Rauhe, T. (2002). Improved Labelling Scheme for Ancestor Queries. In proceedings of the 13th annual ACM-SIAM Symposium on Discrete Algorithm, 2002. 33-43 Cooper, H. (2008). Synthesizing Research: A Guide for Literature Reviews. Harlow: Pearson Education Limited. Creswell, J. (2007). Review of the Literature, Thousand Oaks: Sage Publications. Dellinger, A. B. & Leech, N. L. (2007). "Toward a Unified Validation Framework in Mixed Methods Research" Journal of Mixed Methods Research, Vol. 1, No. 4, pp. 309–332. Diriwächter, R. & Valsiner, J. (2006). “Qualitative Developmental Research Methods in Their Historical and Epistemological Contexts” FQS. Vol 7, No. 1, p. 8 Given, L. M. (2008). The Sage encyclopedia of qualitative research methods. Los Angeles, Calif.: Sage Publications. Green, B. N., Johnson, C. D., and Adams, A. (2006) "Writing Narrative Literature Reviews for Peer-Reviewed Journals: Secrets of the Trade" Journal of Chiropractic Medicine, Vol. 5 No. 3, pp. 101–114. Hart, C. (2008). Literature Reviewing and Argumentation. United Kingdom. London: Sage Hunter, L. & Leahey, E. (2008). "Collaborative Research in Sociology: Trends and Contributing Factors". The American Sociologist, Vol. 39 No. 4, pp. 290. Kasim, R., Alexander, K. & Hudson, J. (2010). A choice of research strategy for identifying community-based action skill requirements in the process of delivering housing market renewal. Research Institute for the Built and Human Environment, University of Salford, UK. Moballeghi, M. & Moghaddam, G.G. (2008). "How Do We Measure Use of Scientific Journals? A Note on Research Methodologies". Scientometrics, 1, pp. 125–133. Remenyi, D., Williams, B., Money, A. and Swartz, E. (2008). Doing Research in Business and Management: An Introduction to Process and Method. London: Sage Robson, C. (2012). Real World Research. Oxford: Blackwell. Sapsford, R. and Jupp, V. (2012). Data Collection and Analysis. London: Sage Saunders, M., Lewis, P. and Thornhill, A. (2003). Research Methods for Business Students. Harlow: Pearson Education Limited. Read More