In the creative environment where research takes place not everything can be improved. The creative “essence” of research must be undisturbed while “accident” wasted effort is reduced to a minimum. In this paper we discuss the types of knowledge at play in the research environment, introduce a new abstract model of knowledge, and using the model explain how we should focus our effort on research students and on particular types of knowledge transfer in order to gain an over all improvement in our research processes. Just as we teach to facilitate student learning, so too can we supervise, teach and guide to facilitate better and faster researching in our academic
computer science departments.
With increased pressure on the research community a way is needed to improve efficiency and productivity. Researchers in Computer Science are in a prime position since research approaches to date have been ad hoc and thus finding improvement (any improvement) should be a simple matter. To cater for the wide breadth of work in the discipline we believe a high level abstraction of the problem and similarly high level
meta solutions are needed at least initially. There will always be new research students constructing their approach to research for the first time. We must gain new insight into the research process and into the way researchers develop their approaches. With this knowledge we can facilitate improved research for them and greater and faster advances for the field.
Before we begin we must understand what we mean by research and what we consider an improvement in it. Improving research efficiency is an optimization problem
under certain constraints. Should your definition of research or your measure of improvement differ from ours the approach we recommend may not be for you.
This paper speaks of research set within certain boundaries. Research is “original investigation undertaken in order to gain knowledge and understanding”  according to the UK’s RAE 2001 guidelines designed to measure research and allocate funding accordingly. Put another way it is “creative work undertaken on a systematic basis in order to increase the stock of knowledge” according to an OECD definition and perhaps
more importantly to the Higher Education Statistics Agency who quoted it [2, 3]. Computer Science Research is taken to mean research that focuses on the creation, adaptation or analysis of computer systems. We limit ourselves to the typical
projects in the university research environment, those involving a student and a supervisor, with occasional input from colleges working in the same area. The typical computer science research project is an idea subject for this discussion as it is
effectively a task of pure knowledge in various forms, from creative ideas, to instantiations of algorithms in code and advances in our understanding of the world through experimental results. The lessons drawn may well apply to a far wider areas of research, but we have not systematically examined this.
To begin, we divide the knowledge needed for research in our field (and presumably all others) into three types. All are required for research, yet only two of these types may be realistically improved. By creating this separation we invalidate the notion that research as a whole cannot be improved because it is all a creative process.
Having isolated the types of knowledge we can target for improvement, we discuss how knowledge can be abstractly measured. We present a model, drawing from it concrete ideas on improving efficiency, and explaining why current software engineering fails to meet the requirements. Finally we discuss one possible way of taking these ideas from abstraction to reality and implementing them in a real research environment.
II. TYPES OF KNOWLEDGE
Lord Kelvin  in 1883 said that “When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it… it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science”. We present here three types of knowledge followed by a conceptual model of research as an individual’s path to new knowledge.
As we try to model something as ethereal as “knowledge” we may not be at an advanced stage of science, but we have at least advanced to the beginning of knowledge, a definition, a model and an encompassing context for further work. Our work relates old concepts in philosophy (stretching back to biblical times) to the modern Information Science and Knowledge Management domains, and finally applies these concepts to the problem of research which is at its core no more than the pursuit of knowledge. To improve the research processes capacity for knowledge generation it is vital we have an understanding of the various types of knowledge, who has them, how they can be created and transferred, and from this we can see how to aid the transfer and facilitate the generation of relevant knowledge while not expending effort to capture data that is superfluous to researchers needs.
In the old testament three types of knowledge are mentioned, Wisdom “chochma”, Insight “bina” and Knowledge “daat”. Rav Kook, an eminent Jewish scholar and philosopher explained the distinction saying that Wisdom provides the underlying framework that allows understanding. Insight is a vision of the future and how things may fit together, the Hebrew word Bina is related to the Hebrew word “boneh”
meaning “to build”. Knowledge (Daat) means a complete attention to detail, i.e. to the facts of the current issue . This is summarized in Table 1. While this is not the only scheme for understanding and partitioning knowledge, it does provide a well establish and useful starting point for discussion.
Read the full paper.
Full citation: Andre Oboler, Simon Lock, Ian Sommerville, “Targetted Improvements,” in proceedings of the second International Conference on Software Engineering Advances (ICSEA 2007), Cap Esterel, France, 25-31 Aug. 2007, IEEE Press, 2007