The information society, having a growing information domain/amount, requires a reliable information infrastructure with ability to distribute the right information from the right information producer to the right information consumer, at the right time and with the right quality on demand (information-on-demand). Another growing demand, parallel with the information distribution infrastructure, is information visualization. Information visualization is one of the important tools that help consumers to interpret and understand the distributed data, and facilitate the decision-making process. Maps have always been a popular method for visualizing spatial objects, objects that contain spatial references such as address co-ordinates and building ID. The rapidly growing need for geographical information systems (also called GI systems/applications) and components has an impact on the processes of development and maintenance of geographical information systems and components. The main purpose of this research is to study design patterns and suggest methods, models, rules of thumb and so on to improve the process of development and maintaining geographical information systems and reduce the total life cycle cost. The main objective is divided into following sub objectives, namely, a) to describe different patterns, solutions, models, suggestions and recommendations that can improve the process of developing GI-applications in order to reduce the cost of producing, developing and maintaining applications with a high level of quality and serviceability, b) to examine if and how the idea of design pattern is applicable to the architecture of software with spatial components (commonly called GI systems), c) to identify the factors that impact the costs of GI software development and maintenance processes and if possible recommend a model for reducing the cost and improving the quality of the solutions, d) to find and suggest solutions, models, rules of thumb, suggestions and recommendations that can improve the process of developing GI-applications. During the study, it has been noticed that recurrent GI-problems are solved but those solutions are hardly communicated among developers. On the basis of software developed in many of the research projects and also students' works we have observed that the architecture of GI software is in most cases similar to any other software. This similarity makes it possible to apply general software development patterns to GI applications. Further more it is noticed that recurring problems are not dependent on who the developer is or to what category she or he belongs, but on the problem situation and the context of the application. This means that the developer faces almost the same kinds of problems in a given situation. It has also been noticed that the size of projects and demands from customers, management, development organisations and culture/experiences, maintenance organisations and so on have an effect on how much the developer group need to use patterns and semi-manufactured components during the software development process. The study has shown that generic IT patterns can be used for designing GI-applications, but there is still a need for technology- and product-specific patterns for GI- components
Luleå: Luleå tekniska universitet, 2005. , 78 p.