Supporting the Collaborative Collection of User's Requirements

Most of the software systems development processes carry high risks for technological companies and service organizations. Some of these risks may ultimately produce delayed or failed projects with low quality software products. The Standish Group in 1999 presented the “Chaos Report”, which determined through a large scale scan on IT projects in northern hemisphere corporations that the software development industry continued to be marked by low quality products and services. According to this report, the main causes of failed projects during 1994 were the following: incomplete requirements (13,1%), little users participation (12,4%), lack of resources (10,6%), non-realistic expectations (9,9%), insufficient executive support (12.4%), changes in requirements and specifications (8,7%), lack of planning (8,1%), little administration of information technology (6,2%), technological incomprehension (4,3%), others (9,9%). It is important to note that a high percentage of errors are related to definition and administration of requirements and little user’s participation. Even though this study was performed some years ago, most of the current software projects are still dealing with the same risks. It is clear that any improvement in requirement engineering may substantially reduce the high risk that menace the quality of decision taken during the negotiations of the user’s final requirements. Requirement engineering includes the following steps (see e.g. Nuseibeh and Easterbrook, 2000): • Collect requirements: identify the stakeholders and obtain their requirements. • Model and analyze the requirements: build abstract descriptions for a better interpretation. • Communicate the requirements: transfer the knowledge among the stakeholders. • Come to an agreement on the requirements: negotiate the requirements with the stakeholders. • Evolve the requirements: manage changes not only in the requirements but also in the environment, or changes generated by the stakeholders.