Vol.3 No.1 2010

74/110

Research paper : A field-scientific approach to Clinico-Informatics (Y. Kinoshita et al.)−71−Synthesiology - English edition Vol.3 No.1 (2010) introduce an essential and effective model of the system by use of the notion of simulation relation in this case. This is an example that the knowledge in basic science worked well enabling such an effective model to be obtained because the research team had known the notion of simulation relation well before this industrial collaboration.The processes of our technology transfer fieldwork were conducted almost precisely according to the description in chapter 4. The element techniques used in this procedure are explained below. Section 5.2 is an explanation of the interview technology used in the first step. Section 5.3 discusses the participant observation used widely in fieldwork. Finally in section 5.4, the model experiment, predevelopment experiment, post-development experiment, restoration experiment, and blind experiment used in trial experiments are described.5.2 InterviewAt the start of fieldwork in technology transfer, the research team must learn the domain knowledge of the field. In most cases, some written material about the domain knowledge is available, and the research team can ask questions, to which the industrial partner answers. We call this step the interview. Since the knowledge and the background or culture of the research team and the industrial partner differ, one must choose the vocabulary carefully during the interview. Yet the communication between the research team and the industrial partner often is so hard at first that it looks almost impossible. This is why we say technology transfer is a cross-cultural exchange. The techniques of interview overlap in a large part with requirement analysis in system and software engineering.Interview is done during the whole process of fieldwork, not just at the start. So, the project should be scheduled taking the time for interview into account. Another thing to consider is that the interviewee is not necessarily cooperative to the interview, especially when he/she is not a member of the project, as the interview must occasionally be done with a person who is not a project member. Special preparation for interviews should be considered for such cases.5.3 Participant observationParticipant observation is one of the standard techniques in qualitative research. The observer becomes a part of what is observed, therefore the act of observation itself may effect the result of observation. It is widely used in ethnology and sociology.Processes of information system design and development are typical examples of non-reproducible processes. If the methodology of experimental science is applied simple-mindedly, one would immediately face with difficulties caused by the non-reproducibility, as well as the influence of the observer on the object.5.4 Trial experimentThere are several objectives in trial experiments.1) To give examples of application of the technology of concern in the context of the industrial partner, in order to make it easier for them to evaluate the value of the technology.2) To exemplify applications of the technology of concern.3) To ease the learning of the technology by the industrial partner, by means of tutoring by the research team.Of these, 1) and 3) are for the interest of the industrial partner, while 2) is for that of the research team. No special method is necessary for 2) and 3), but we make a list of some specific methods for 1) (Fig. 2).a) [Model experiment] Trial experiment for past prototypes and past development cases The risk is small since there is no damage even if the technology application fails.b) [Reconstruction experiment] Take an old system development, for a part of which the recorded document is missing. Reconstruction experiment is conducted by tracing exactly the old development according to the recorded document wherever it is available and making up the undocumented part by doing new development. For example, the specification is rebuilt in consultation with the preserved source code if the specification document is missing, and model checking is done according to the newly built specification document and the source code. c) [Pre-development experiment] Trial experiment for a commercial product under development or for its development processd) [Post-development experiment] Trial experiment for a commercial of products already available in the market This includes conducting analysis of failures reported Engineers onlyScientistsand engineersScientists onlyInvolved membersExperiment ofapplication of technologyForm ofexperimentBlindPost-devel.Pre-devel.Reconstr.ModelFig. 2 Classification of experiments of application of technology.

※このページを正しく表示するにはFlashPlayer9以上が必要です