Decision model for information system development process adaptation to individual needs of the projects

This dissertation presents formal definition of the decision model for information systems development process adaptation to individual needs of the projects. The ability of the software process to adapt to project specific needs, presents one of the most important factors for assuring methodology suitability and acceptance among software engineers in particular organizational system. The idea of adaptation of information systems development methodology is not new. Inability of traditional methodologies to adapt to particular situational needs, their rigidness and technical unsuitability led to development of special scientific field, called method engineering. The field of method engineering is very complex, because it focuses mainly on theoretical point of view of methodology construction and adaptation, and it uses very complex mathematical formalisms. That is also the main reason why method engineering has never been widely acknowledged or practiced by software engineers. Method engineering also doesn't deal with social suitability and acceptance of the methodology among its users. That has been another reason why the field hasn't been accepted among the engineers. With the emergence of the new approach, also called "agile approach", the idea of systems development methodology adaptation seems to become more attractive, as this time the initiative comes from practioners. The idea of agile approaches is very similar to the idea of method engineering, only that agile approaches are not so complex. In contrast to method engineering the focus in agile methodologies is on the adaptation of the process. Moreover, agile methodologies seem to give more attention to sociological aspects of the methodology, which have been rather ignored in method engineering. The key to successful introduction of systems development methodologies into organizations, their later use and acceptance, lies in the ability of its prescribed software process to be adaptable to project specific needs. The methodology has to define flexible software process, capable of adapting to specific situations, in which new projects emerge. The process adaptation has to be supported by software tool which automates the adaptation procedure and facilitates the work of method engineer. The aim of this dissertation is to combine theoretical knowledge from method engineering and practical knowledge from organizations for software development. On this basis we formally define decision model for information systems development process adaptation to individual needs of the projects. The decision model serves as a basis for definition of the specification for software tool prototype. The approach to process adaptation presented in this dissertation is based upon selecting the most suitable path through the given process, used in the organization. The information system development process used in the organization is defined as a base process. The approach extends the idea of choosing one of the possible paths through the given methodology, defined in situational method engineering field. Our extension of the approach suggests abolition of predefined paths. The construction of adapted instance of a base process is directed by decision model, on the basis of decision rules. For every process element, the decision model has to give an answer about its inclusion in adapted instance of the base software process. Decision rules are used for construction of adapted instances, assuring integrity of instances and definition of dependencies among project characteristics. In this dissertation we present decision rule metamodel and structure of proposed decision rules. Project characteristics are used to describe projects. Beside technological characteristics, we have also defined sociological characteristics and customer requirements in this dissertation. Multiple types of characteristics ensure, that the adapted instance of the base process meets various technological and sociological needs of the project. Formal definition of the decision model along with decision rules establish theoretical basis for the software tool prototype specification. The software tool is intended to help method engineer at his work. We used use cases to describe expected functionality of the software tool. The tool defined in the dissertation represents complete solution for process adaptation to project specific needs, because it considers technological and sociological point of view about base methodology use, used in the particular organization