No description available.

Context: With the rising popularity of machine learning, looking at its shortcomings is valuable in seeing how well machine learning is applicable. Is it possible to apply the clustering with a small dataset? Objectives: This thesis consists of a literature study, a survey and an experiment. It investigates how two different unsupervised machine learning algorithms DBSCAN(Density-Based Spatial Clustering of Applications with Noise) and K-means run on a dataset gathered from a survey. Methods: Making a survey where we can see statistically what most people chose and apply clustering with the data from the survey to confirm if the clustering has the same patterns as what people have picked statistically. Results: It was possible to identify patterns with clustering algorithms using a small dataset. The literature studies show examples that both algorithms have been used successfully. Conclusions: It's possible to see patterns using DBSCAN and K-means on a small dataset. The size of the dataset is not necessarily the only aspect to take into consideration, feature and parameter selection are both important as well since the algorithms need to be tuned and customized to the data.

Background: Software quality assurance (QA) is an essential activity in the software development lifecycle. Among the different QA activities, regression testing is a challenging task for large-scale software development. Regression testing is a well-researched area, and a large number of techniques have been proposed to fulfill the needs of industry. Despite the extensive research, the adoption of proposed regression testing techniques in the industry is limited. Studies show that there is a visible gap between research and practice. Objective: This work aims at reducing the gap between industry and academia in regression testing. To fulfill this aim we have the following objectives: 1) Understanding the practitioners' goals regarding regression testing. 2) Understanding the current state of regression testing practice and challenges in the industry. 3) Investigating the testing research applicable in an industrial context. Method: We conducted multiple studies using different methods. To explore the industry perspective on regression testing we used focus group and interview-based studies. To explore solutions from the literature, we used the systematic literature review and systematic mapping study. Results: This thesis presents the practitioners' specific regression testing goals. The identified goals are confidence, controlled fault slippage, effectiveness, efficiency, and customer satisfaction. The challenges identified in the thesis are of two categories, 1) management related challenges and 2) technical challenges. Technical challenges relate to test suite maintenance, test case selection, test case prioritization, evaluation of regression testing. We have mapped 26 empirically evaluated regression testing techniques to the context, effect, and information taxonomies, and provided a guide to the practitioners regarding the adoption of the techniques in an industrial setting. We have also classified 56 model-based test case generation techniques regarding their strengths/limitations, input/intermediate models used, and relevance to the industrial context. Conclusions: The challenges identified in this study are not new for research and practice. There could be two reasons regarding the presence of recurring challenges: 1) regression testing techniques proposed in the literature do not fit the companies’ context, 2) or, companies are not aware of the availability of the techniques that could be suitable for their context. To support the adoption of existing research on regression testing in the industry, we have presented three taxonomies. These taxonomies, allow the characterization of regression testing techniques and enable to determine which of these techniques might be suitable in a given context. Furthermore, the identification of information needs for these techniques would be helpful to learn the implications regarding the cost of adoption. Regarding the support in test case generation, we conclude that current research on interaction model-based test case generation techniques did not illustrate the use of rigorous methodology, and currently, model-based test case generation techniques have low relevance for the industrial problems.

In today’s society, we tend to use our mobiles more and more and they have become a importent part in our everyday life. Users have become more aware of what is a good application and creating an appealing application with high usability becomes of great importance. This study has investigated how it is possible to create a useful application based on predetermined criteria and design patterns. The application was implemented in React Native with Firebase as a backend. The design process was carried out with user tests as well as think aloud as the basic method. The results were evaluated and analyzed, the shortcomings considered to affect the usability were corrected. To assess the user-friendliness of the application, a System Usability Scale, SUS evaluation was conducted. The result shows how important it is that the implementation is preceded by a good design process to make the final product as good as possible. In this study, the result became an application that the users of the product graded relatively high in the SUS evaluation.

This report describes work done on implementing automatic detection of looping frame sequences in a video signal. The central loop detection algorithm is described. Hashing of video frames is used as a means of improving computational performance. Two video signals are compared with respect to containing loops, and switching of displayed stream is done based on evaluated stream qualities. Repeating sequences -distinct from looping sequences- are also discussed, as well as cursory thoughts for further work on implementing a more comprehensive error detection package for video signals. / Denna rapport beskriver arbete utfört med att implementera automatisk detektion av loopande följder av bildrutor i en videosignal. Den centrala algoritmen för igenkänning av loopar beskrivs. För att förbättra beräkningshastighet tillämpas hashning av bildrutor. Två videosignaler jämförs med avseende på deras innehåller av loopar, och växling av uppspelad signal sker baserat på utvärdering av signalernas kvalitet. Upprepade följder, skilt från loopande följder, diskuteras. Tankar för vidare utveckling av en mer heltäckande paketlösning för detektion av olika typer av fel i videosignaler presenteras kort.

The purpose of this study is to examine and compare a small set of software stacks which can be used for a progressive web application to see if one stack is more suitable to build a journaling system with strict requirements on accessibility and traceability. The study is a comparative case study where the development of a system with different software stacks were compared. First a literature study was conducted to find potential technologies to use. Then interviews were conducted with persons who have used journaling systems within their work to get a clear picture of the essential functionality of a journaling system. The purpose of this was to create a specification of a simple journaling system. This specification was used during the development of two software stacks. The literature study showed that two databases and two javascript-framework of the researched potential technologies were more suitable than the other technologies. These technologies were then used in the two software stacks where code size and development time was measured. In the end there was not a big difference between the two software stacks in terms of code size and development time. Though one of the stacks had a small advantage of having a marginally faster development time and smaller code size. The conclusion was that a more comprehensive study should be done to get a more conclusive answer but that the results of this study can be used together with other data as a decision guide when a software stack is chosen for a project for a progressive web application with strict requirements on accessibility and traceability. A limitation had been set at only choosing two software stacks to compare because the lack of time. Also, the number of features in the finished systems has been limited, because of the lack of time only the essential parts of a journaling system will be included. The study also doesn’t present or take into account all the laws and requirements that affects a journaling system

Background: In Agile Software Development (ASD) planning is valued more than the resulting plans. Planning and estimation are carried out at multiple levels in ASD. Agile plans and estimates are frequently updated to reflect the current situation. It supports shorter release cycles and flexibility to incorporate changing market and customer needs. Many empirical studies have been conducted to investigate effort estimation in ASD. However, the evidence on effort estimation in ASD has not been aggregated and organized. Objective: This thesis has two main objectives: First, to identify and aggregate evidence, from both literature and industry, on effort estimation in ASD. Second, to support research and practice on effort estimation in ASD by organizing the identified knowledge. Method: In this thesis we conducted a Systematic Literature Review (SLR), a systematic mapping study, a questionnaire based industrial survey and an interview based survey. Results: The SLR and survey results showed that agile teams estimate effort, mostly during release and iteration planning, using techniques that are based on experts' subjective assessments. During effort estimation team related cost drivers, such as team members’ expertise, are considered important. The results also highlighted that implementation and testing are the only activities that are accounted for in effort estimates by most agile teams. Our mapping study identified that taxonomies in SE are mostly designed and presented in an ad-hoc manner. To fill this gap we updated an existing method to design taxonomies in a systematic way. The method is then used to design taxonomy on effort estimation in ASD using the evidence identified in our SLR and survey as input. Conclusions: The proposed taxonomy is evaluated by characterizing effort estimation cases of selected agile projects reported in literature. The evaluation found that the reporting of the selected studies lacks information related to the context and predictors used during effort estimation in ASD. The taxonomy can be used in consistently reporting effort estimation studies in ASD to facilitate identification, aggregation and analysis of the evidence. The proposed taxonomy was also used to characterize the effort estimation activity of agile teams in three different software companies. The proposed taxonomy was found to be useful by interviewed agile practitioners in documenting important effort estimation related knowledge, which otherwise remain tacit in most cases.

No description available.

Context: In order to provide more functionalities and services, systems collaborate with each other creating more complex systems called Systems of Systems. Exploiting security vulnerabilities in such complex systems has an impact over system safety and it is not sufficient to analyze them separately in the development process. Observing these safety and security interdependencies together can be done via the process of attack modeling where attack models deploy a model for detecting vulnerabilities and possible mitigation strategies while observing system security from an adversary point of view. Objective: The aim of the thesis is to explore the interdependencies between safety and security concerns, to provide details on attack model(s) and affected safety requirements of given system, to argument that the system is acceptably safe to operate, and to contribute to the identified business challenges. Method: The thesis work consists of: (i) a literature survey on interdependencies between safety and security, and a literature survey on the existing attack models; (ii) a demonstration on a use case where the argument that the given system is acceptably safe with respect to the selected attack model has been provided using Goal Structuring Notation (GSN). Conclusion: The first literature survey conducted on the topic of interdependencies between safety and security has resulted in a number of papers addressing the importance of investigating safety and security together. Reviewed papers have been focused either on proposing new approaches or extending the existing ones in different industry domains like automotive, railway, industrial, etc. The literature survey on existing attack models has resulted in a number of papers elaborating attack models in general and showing domain-specific attack models such as those in control systems, vehicles, Cloud Computing, IoT, networks, RFID, Recommender Systems, etc. To provide an argument that the given system is acceptably safe by using GSN, investigated results from the selected attack model showed how to protect system while observing it from an adversary point of view. Including security countermeasures, i.e. data and identity authentication and implementation of access control in the system development process can produce an acceptably safe system, whilst, at the same time, affect different business aspects by introducing latency and delay to the system. However, avoiding such mitigation techniques may have catastrophic impact on the system and its environment when attacks are launched.

The Internet of Things is a growing market that creates opportunities and chal- lenges for companies. Bergeforsen Obstacle Race is an obstacle race which 2016 attracted 250 people to participate. To investigate new ways of creating a more giving event for competitors and spectators, an independent work has dur- ing its 10 weeks done a theoretical study of Internet of Things with its popular technology Radio Frequency Identification, RESTful web services and Digest authentication to lay the grounds for a RESTful web service which purpose is to enable modules of user interfaces. Impact mapping and user stories was used before construction to build a product backlog. During construction agile meth- ods from Cowboy was followed. The service was developed with modern tech- nology like Node.js and MongoDB. An overview of a scenario of the event has been used to visualize how it could be used where nodes, which are computers with RFID reader modules, can read a competitor and send their identifier with time to a server through RESTful methods. To store sent data, database objects have been created for races, competitors and readings. The security of the sys- tem has been investigated by implementing Digest authentication and database Role Base Access Control. The result is a RESTful web service with 2 routes to modify races and competitors and 2 more routes to give functionality to adding and removing competitors from races and readings. The routes have been con- trolled with functional testing, and their value investigated. At last a comparison have been made with the current used system and the constructed RESTful web service. The independent work was then examined to show that a closer cus- tomer relationship could lay the foundation for a better system and comparison but motivated a RESTful web service as there was only specifications of cus- tomer values. The work has provided ethical aspects regarding licenses, security and sensitive information. / Sakernas Internet är en växande marknad som skapar nya möjligheter och ut- maningar hos företag. Bergeforsen Obstacle Race är en hinderbanetävling som 2016 drog 250 personer till att delta. För att undersöka nya sätt att kunna skapa ett roligare evenemang för deltagare och åskådare har det självständiga arbetet som genomförts under 10 veckor genomfört en teoristudie av Sakernas Internet med populära tekniken Radiofrekvensidentifiering, REST-baserade webbtjäns- ter och Digest autentisering använts för att ligga till grunden för en REST-base- rad webbtjänst som ska kunna möjliggöra moduler av användargränssnitt. Innan konstruktion har Impact Mapping och användarhistorier använts för att bygga en product backlog. Under konstruktion har agila metoder från utvecklingsme- toden Cowboy följts för att färdigställa systemet. Systemet har utvecklats med moderna tekniker som Node.js och MongoDB. I systemet har en överblick av ett evenemang skapats för att visualisera hur systemet skulle kunna användas. I det finns de noder som med RFID-modul ska kunna läsa av en deltagare och skicka det till en server genom REST-metoder. För att kunna lagra den data som skickas har databasobjekt skapats för typerna evenemang (races), deltagare (competitors) och avläsningar (readings). Tjänstens säkerhet har tagits i åtanke då Digest autentisering och rollbaserad åtkomstkontroll till databas implemente- rats. Resultatet har blivit ett system som i öppen källkod möjliggjort modifie- ring av resurserna races och competitors genom två routes. Utöver det har ytter- ligare två routes skapats för enklare modifiering av tillägg och borttagning för deltagare av evenemang och avläsningar. För att säkerställa systemet har funk- tionella tester genomförts och mervärdet undersökts. Till sist har en jämförelse mot nuvarande bakomliggande system genomförts. Det självständiga arbetet granskades sedan för att visa på att en närmare kundrelation kunde legat till grunden för ett bättre system och jämförelse. Eftersom att det endast fanns spe- kulationer av mervärden så motiverades det REST-baserade systemet. Arbetet har gett etiska aspekter gällande licens, säkerhet och känslig information.