<p>The toolkit has been implemented as planned: The ground work for visualisation mappings and relationships between datasets have been finished. Wavelet transforms have been to compress datasets in order to reduce computational time. Principal Component Analysis and other transforms are working. Examples of use have been provided, and several ways of visualizing them have been provided. Multivariate Image Analysis is viable on regular Workstations.</p>

<p>Role based access control (RBAC) is widely used in health care systems today. Some of the biggest systems in use at Norwegian hospitals utilizes role based integration. The basic concept of RBAC is that users are assigned to roles, permissions are assigned to roles and users acquire permissions by being members of roles. An alternative approach to the role based access distribution, is that information should be available only to those who are taking active part in a patient’s treatment. This approach is called decision based access control (DBAC). While some RBAC implementations grant access to a groups of people by ward, DBAC ensures that access to relevant parts of the patient’s medical record is given for treatment purposes regardless of which department the health care worker belongs to. Until now the granularity which the legal framework describes has been difficult to follow. The practical approach has been to grant access to entire wards or organizational units in which the patient currently resides. Due to the protection of personal privacy, it is not acceptable that any medical record is available to every clinician at all times. The most important reason to implement DBAC where RBAC exists today, is to get an access control model that is more dynamic. The users should have the access they need to perform their job at all times, but not more access than needed. With RBAC, practice has shown that it is very hard to make dynamic access rules when properties such as time and tasks of an employee’s work change. This study reveals that pretty much all security measures in the RBAC systems can be overridden by the use of emergency access features. These features are used extensively in everyday work at the hospitals, and thereby creates a security risk. At the same time conformance with the legal framework is not maintained. Two scenarios are simulated in a fictional RBAC and DBAC environment in this report. The results of the simulation show that a complete audit of the logs containing access right enhancements in the RBAC environment is unfeasible at a large hospital, and even checking a few percent of the entries is also a very large job. Changing from RBAC to DBAC would probably affect this situation to the better. Some economical advantages are also pointed out. If a change is made, a considerable amount of time that is used by health care workers to unblock access to information they need in their everyday work will be saved.</p>

<p>Software is gradually replacing paper based administration systems. The migration to electronic systems is supposed to make life easier for the users. If this is to be the case then these software systems must be created in such a way that the end users are able to use them effectively. To achieve usable systems, software testing must be utilized. There are many ways to test a program, with or without involving real users. Scenario testing is a somewhat poorly documented discipline in software testing, with ambiguous definitions. It does however seem to be well suited in combination with users to test external parts of a software system in a late state of development. This project is based on the work done in the software engineering depth study [12]. There we conducted empirical work and internal testing of the software system KAS, and laid the foundation for this Master’s thesis. In this report we have continued the work with this software and concentrated on its external characteristics and user testing. We have analyzed scenario testing further through a software test of this system involving its future users. The users have been given tasks to complete through stories that explain what to do but not how to do it. We have observed the test subjects closely throughout the tests, and collected important data. The results have been evaluated in order to assess their usefulness, which further points to the quality of scenario testing as a testing method. The results have also spawned functional requirements which have been implemented into the KAS. Through this project we have gained experience that can be useful to others conducting scenario tests or doing research in software testing in the future.</p>

<p>This document investigates how real-time algorithmic music composition software constrains and shapes software architecture. To accomplish this, we have employed a method known as Action Research on the software system ImproSculpt. ImproSculpt is real-time algorithmic music composition system for use in both live performances and studio contexts, created by Øyvind Brandtsegg. Our role was to improve the software architecture of ImproSculpt, while gathering data for our research goal. To get an overview of architecture and architectural tactics we could use to improve the structure of the system, a literature study was first conducted on this subject. A design phase followed, where the old architecture was analyzed, and a suggestion for a new system architecture was proposed. After the design phase was completed, we performed four iterations of the action resesarch cyclical process model, where we implemented our new architecture step by step, evaluating and learning from the process as we went along. This project is a follow up of our previous research project, “Artistic Software” [3], that investigated how algorithmic composition was influenced by software.</p>

<p>BUCS Implementing safety An approach as to how to implement safety concernsting safety</p>

<p>Our master thesis is an extension based on our thesis written in the autumn 2005.</p>

<p>Today it has become common practice to apply systematic reuse during software development. By reuse, the gain from creating a piece of software can be multiplied, as instead of creating a new component each time, old ones can be reused. This increases productivity (shorter time-to-market, less cost) and also software quality, as the components get well tested through using them in different systems. There are, however, many ways of applying reuse. There are different types of components that can be applied in systematic reuse. The most common ones are internally developed, OSS, COTS, or outsourced components. There are also many different ways to share and access the components among the developers. Today all companies who apply reuse have some sort of distributed way of sharing. To use product families is also one way of applying reuse. This can take reuse to another level as the reused parts can be vast, but it can also be used for branding a line of products. The main part of this thesis is a quantitative survey that was executed with a questionnaire. 32 Norwegian software companies participated in the survey. The questionnaire asked about who applied reuse and product families, how they applied it, and what the respondents thought were important when applying it. The data collected is used to answer 3 research questions and are also discussed against related research. The data is also used to see if there are any differences between how reuse is applied in companies of different sizes and internally in departments as well as for whole companies. Also the impact of different program languages and development processes/methods on reuse is explored. This survey builds upon the pre-study “Reuse through product-families and framework” [MS00]. In the pre-study subjects from 12 Norwegian software development companies were interviewed about how they utilized reuse and product families. This was a qualitative survey with open questions, which was used to discover trends in Norwegian software development companies, and these trends are in this thesis examined. The data from another survey done by IKT-Norge is also used in this thesis, but only the questions added extra for NTNU. These were about process improvement as well as reuse. There were a total of 142 Norwegian companies that responded, and 60 who answered the extra questions. The IKT-Norge survey is also compared against the thesis survey.</p>

<p>Quality Management in software development is a topic that has become very important. A lot of resources and effort has been invested in making formal routines and process descriptions, which have resulted in extensive systems, so-called Quality Management Systems(QMS). This thesis investigates attitudes towards such systems, seen from both software developers and quality managers points of view. During the fall of 2005, a student project in the subject Depth Study in Software Engineering TDT4735, in form of a qualitative empirical study was performed. Through interviews with developers and quality managers in different Norwegian software companies, it identified some interesting issues to further investigate. As an extension of that depth project, this Masters' Thesis goes further into the area of Quality Management Systems, by presenting a quantitative study. The main research questions for this thesis are: RQ1 Certification today, a must or just more work? RQ2 Developers vs Managers. A battle for quality? RQ3 How make a QMS work? In addition to the main empirical study, the thesis also consists of a brief field study towards two of the largest ICT organisations in Norway, and describes what they consider to be the most interesting and challenging aspects of quality assurance and process improvement in Norwegian software industry.</p>

<p>This thesis presents the results of an explorative survey of industrial involvement in open source. The survey is performed in collaboration with participants from the European ITEA project, COSI (Co-development using inner & Open source in Software Intensive products). The survey was performed to explore aspects of industrial involvement in open source, as industrial management of open source project, industrial use of open source components, related development, and communication processes, and industrial relationships to the open source community. The survey is based on a survey of open source literature, structured interviews of IT industry personnel, and a questionnaire. Three Norwegian companies participated in the interviews, which were used to explore interesting issues, and to provide input to the questionnaire. The questionnaire collected answers from industrial partners in the European COSI project. Our research has an explorative research goal, and five research questions. Each research question shed light on an interesting topic, and their answers will be presented accordingly. These answers are in form of both of qualitative and quantitative answers, but primarily descriptive information. It has been found that companies participate in open source related development with the roles open source owner, open source participant, inner source participant, and user of open source components. These roles will be explained further, together with the answers to the five research questions. We start with a presentation of the motivations behind use of open source components, and the selection processes used to find, and evaluate these components. Then, experiences from companies which have started their own open source projects will be discussed. Next, we describe the use of software development methodologies from open source, within companies. Furthermore, we present results related to the impact open source, have on company internal development processes. Finally, industrial participation in open source projects by companies will be analyzed. The answers to the research questions form part of the contribution of this thesis. Some of these results are used to create guidelines for open source management, and open source component selection. We have also proposed twelve hypotheses, based on our findings. These hypotheses will work as a basis for future studies, together with other results, and our research design. Future studies may include distribution of the questionnaire to a bigger sample, focus on fewer issues, and use of other research methods to shed light on these issues.</p>

<p>Open source is one of the more interesting trends in software engineering today. The goal of the software engineering discipline is to increase efficiency in the development process, and maximize quality of the product. Open source development processes offer the potential for reducing costs for commercial enterprises. This master's thesis addresses how open source documents architecture, and how it uses documentation in general. Open source has a reputation of creating high quality software, but documentation of process and product is weak. This may be a hurdle for wider adoption of open source processes, as a thorough understanding of a product's qualities is central to its success. The goal is to better understand documentation requirements in open source. The study is based on participation to the Apache Derby open source project. Action research is the research method. The findings show that the Apache Derby documents its artifacts in a number of ways, but fails to aggregate it in a meaningful way. A rich set of written communication mediums compensate for this by giving developers the ability to understand the product over time. The study suggests the popularity and diffusion of an open source project may affect requirements for documentation.</p>