MSc Software Engineering Dissertation Finding a Research Problem

MSc Software Engineering Dissertation Finding a Research Problem and Additional Guidance Stewart Green February 22 nd, 2006 Software Engineering Dissertation Guidance

Purpose l This presentation gives guidance on how to: – – – 2 Find a research topic and a research question Structure your “Abstract”, and “Introduction” and “Conclusion” chapters Create and review the organisation of your dissertation report February 22 nd, 2006 Software Engineering Dissertation Guidance

Helpful Books l l l 3 The Craft of Research, Wayne Booth et al. , 2 nd edition, Chicago Press, 2003 Projects in Computing and Information Systems: A Student’s Guide, Christian Dawson, Addison Wesley, 2005 Both in the library, on short-term loan February 22 nd, 2006 Software Engineering Dissertation Guidance

Assessment of the Dissertation: a Reminder l l Research: 50% Software Engineering: 50% – – 4 Creating a computer-based system to address a research question You need to follow the whole software development lifecycle from requirements through to testing and deployment. February 22 nd, 2006 Software Engineering Dissertation Guidance

Finding a Topic l Find a topic specific enough to let you master a reasonable amount of information on it – – 5 NOT “The history of requirements engineering” BUT “Automated support for traceability in requirements engineering” OR “Model-based paradigms in formal specification” OR “Learning to use software patterns” February 22 nd, 2006 Software Engineering Dissertation Guidance

From Topics to Questions l l You need to identify a research question related to your topic How? First: ask questions about your chosen topic, e. g. : – – – 6 What techniques support traceability in requirements engineering? How well does Java support concurrency? What proof systems are available for supporting data refinement in model-based formal specification paradigms? February 22 nd, 2006 Software Engineering Dissertation Guidance

From Topics to Questions l To help to identify good questions, you must research your topic: – In the library l l – – – 7 Look at relevant text books and journals Look at relevant databases (e. g. Inspec) Talk to colleagues and lecturers Review related online discussion groups Search scholar. google. com February 22 nd, 2006 Software Engineering Dissertation Guidance

From Topics to Questions l You need to focus your topic and question(s) – – – 8 Ask question that are worth pondering (e. g. “how” and “why” questions, e. g. : “How can novice programmers learn to use software patterns? ”) What are your topic’s key concepts? How do they interrelate? Questions about effectiveness, efficiency and utility may be productive? (“Can the elicitation of requirements be achieved at a reduced cost using technique x? ”) February 22 nd, 2006 Software Engineering Dissertation Guidance

So What? l l 9 When you think that you have a good question, ask of your question “So what? ” in order to try to identify the wider significance of any answer you find to your question. E. g. : A method for goal-oriented requirements engineering has been synthesised”, but SO WHAT? Who cares? Why should they care? You must find out and tell us in your dissertation report. February 22 nd, 2006 Software Engineering Dissertation Guidance

Test Your Progress l You should be able to write down: – – – l For example: – – – 10 TOPIC: I am working on x, QUESTION: Because I want to find out (who, what, when, why, …) SIGNIFICANCE: In order to help the research community understand (how, why, whether, etc. ) I am working on goal-oriented requirements engineering Because I want to synthesise a goal-oriented method In order to compare the goal-oriented approach to RE to other RE approaches February 22 nd, 2006 Software Engineering Dissertation Guidance

Claims l l l 11 The result of addressing the problem that you have identified with the solution that you are suggesting should lead ultimately to one or more claims, e. g. “Goal-oriented approaches are clearly more efficient, effective and useable than other approaches to requirements elicitation” You will need to backup this claim with evidence. This may be, for example, data produced by your computer-based system, data you discover using your computer-based system, the fact that you have been able to produce a computer-based system of this sort at all… In a research report you make a claim, back it with reasons, based upon your evidence, and acknowledge and respond to others’ views. February 22 nd, 2006 Software Engineering Dissertation Guidance

“Abstract”, and “Introduction” and “Conclusion” Chapters of the Dissertation Report l Introduction chapter: – – Describe and discuss the background context to your work on this topic Describe the problem(s) that motivate your work: l l – Describe how you are addressing the problem l l 12 Both: a lack of knowledge about something And: a cost or penalty associated with that lack of knowledge Or: some benefit for producing that knowledge Either state the outline of your solution Or promise a solution at the end of the dissertation February 22 nd, 2006 Software Engineering Dissertation Guidance

“Abstract” l Your abstract should be a brief summary of the paper, with just one or two (preferably one) sentences on each of the following, in this order: – – – 13 The context of this work The problem Your solution Your result(s) i. e. claim The wider significance of your results (the answer to the “So what? ” question February 22 nd, 2006 Software Engineering Dissertation Guidance

“Conclusion” chapter l l l 14 Assess the extent to which each of your claims has been supported State the wider significance of your results Call for more research (describe in outline the new research that is needed) February 22 nd, 2006 Software Engineering Dissertation Guidance

Creating and Reviewing the Organisation of Your Dissertation l l 15 Start with the overall organisation of your dissertation, i. e. its chapter structure; your supervisor should be able to guide you Then move on to the organisation of each chapter in terms of sections and subsections (again, supervisor can guide) Then think about structure and clarity of each paragraph, and of each sentence in it This will entail thinking about punctuation, which is important (compare “Eats, shoots and leaves” to “Eats shoots and leaves”: do they mean the same thing? Is meaning important? ? ? ) Software Engineering Dissertation February 22 nd, 2006 Guidance