June 13th – June 18th 1999, Dagstuhl Seminar 99241
Requirements Capture, Documentation and Validation
E. Börger (Pisa), B. Hörger (Daimler-Benz, Ulm), D. Parnas (McMaster), D. Rombach (Kaiserslautern)
For support, please contact
The aim of the seminar is to bring together software engineering researchers from academia and software engineers from industry to critically compare the practice and research for capturing, documenting and validating software requirements. The focus will be on the industrial strength of the used methods and on their relevance for the production of large software. The seminar will be guided by a casestudy [http://www.iese.fhg.de/Dagstuhl/seminar99241.html] (formulated by the Fraunhofer Institut for Experimental Software Engineering in Kaiserslautern) and is proposed to the participants for an illustration of their contribution to the seminar. We aim at achieving a concrete evaluation which helps to understand the reasons for the gap between current practice and much of current research and to bring the two closer to each other.
In order to capture the full range of problems which are encountered in industrial applications, in addition to the requirements specification document also the process of capturing, documenting and validating requirements in the general sense of the term has to be considered. This includes the early, mostly informal phases of capturing requirements which come in the form of typically non-functional conditions (e.g., quality requests). Methods for these early phases have mainly been developed in the context of information systems. The transfer of such methods to embedded systems which are the focus of the case study will be an interesting area of discussion for the seminar.
In particular the seminar shall cover methods for elicitation, specification, inspection, testing and verification of requirements.
Typical questions to be answered in the context of these methods are:
- can they foster a good understanding of requirements (and a reliable capture of the users' statements of their requirements)?
- can they help to explore whether the requirements really do describe a system which meets the users' needs?
- can they contribute to practical inspections, which help to detect errors at the earliest possible stage of software production - well known to be responsible for most of the costly errors occurring during the software development process ?
- can they support good documentation, their reusability, modifiability? etc.
The organizers do not want to exclude any particular approach or method and therefore solicit also contributions which try to adopt so-called formal techniques, but with the specific goal of exposing them to the question of their practicality for industrial usage.
In accordance with the goal to evaluate scientific methods and to compare them to current industrial practice, the list of invited persons tries to balance the participation from industry and from research and to represent the main streams in current research.
Participants experienced in a (specification) method are invited to apply it to the informal requirements of the casestudy [http://www.iese.fhg.de/Dagstuhl/seminar99241.html].
The other participants are invited to present talks highlighting specific problems and their solution in the area of requirements capture, documentation and validation for embedded systems using the case study for illustrative purposes. These contributions will be compared and summarized through evening working groups and evaluation sessions. For efficient organization of the seminar all participants are required to submit to the organizers, a short (not more than one page) description of the scope of their contribution. Please send your abstract to firstname.lastname@example.org by the end of January.
Depending upon the outcome of the seminar the organizers plan to collect revised and then reviewed versions of the contributions into a publication.