The increasing number of IoT devices used in today’s society has led to a demand for better security in order to prevent attackers from gaining access to private information. The IoT brings a wide application scope and because of that there are a lot of ways to set up a secure network and manage keys in these kinds of networks. This paper presents a comparison between the security model in Zigbee and the new recommended practice for Key Management Protocols defined by the IEEE 802.15.9 standard. We investigate key establishment and transportation together with the vulnerabilities that this might bring regarding potential attacks like DoS and MitM. Since these protocols are built on the IEEE 802.15.4 standard, experimental tests have been made where we analyze the throughput, RTT and packet loss over varied distances and we try to determine the maximum transmission range for devices using IEEE 802.15.4 modules. The IEEE 802.15.9 standard works with different KMPs and depending on the KMP being used we can see both similarities and differences regarding key management and possible attacks when comparing it to ZigBee. Furthermore, we found that attacks on a ZigBee device is more likely to compromise the whole network while similar attacks would only affect the specific peers in an IEEE 802.15.9 communication. Based on the experiments we find that open areas, distance and interference have a negative effect on the throughput, RTT and packet loss of the communication.

Continuous authentication using behavioral biometrics can provide an additional layer of protection against online account hijacking and fraud. Mouse dynamics classification is the concept of determining the authenticity of a user through the use of machine learning algorithms on mouse movement data. This thesis investigates the viability of state of the art deep learning technologies in mouse dynamics classification by designing convolutional neural network classifiers taking mouse movement images as input. For purposes of comparison, classifiers using the random forest algorithm and engineered features inspired by related works are implemented and tested on the same data set as the neural network classifier. A technique for lowering bias toward the on-screen location of mouse movement images is introduced, although its effectiveness is questionable and requires further research to thoroughly investigate. This technique was named 'centering', and is used for the deep learning-based classification methods alongside images not using the technique. The neural network classifiers yielded single action classification accuracies of 66% for centering, and 78% for non-centering. The random forest classifiers achieved the average accuracy of 79% for single action classification, which is very close to the results of other studies using similar methods. In addition to single action classification, a set based classification is made. This is the method most suitable for implementation in an actual authentication system as the accuracy is much higher. The neural network and random forest classifiers have different strengths. The neural network is proficient at classifying mouse actions that are of similar appearance in terms of length, location, and curvature. The random forest classifiers seem to be more consistent in these regards, although the accuracy deteriorates for especially long actions. As the different classification methods in this study have different strengths and weaknesses, a composite classification experiment was made where the output was determined by the least ambiguous output of the two models. This composite classification had an accuracy of 83%, meaning it outperformed both the individual models.

I det här arbetet har en enkätundersökning genomförts för att undersöka de förutsättningar lärare haft vid införandet av programmering i läroplanen samt hur lärare upplevt förberedelserna inför förändringen. Enkäten besvarades av verksamma mellanstadielärare i en medelstor stad i södra Sverige. Datamaterialet analyserades därefter med utgångspunkt från tidigare forskning. Studiens resultat visar att den största andelen lärare som utbildat sig inom programmering har gjort det på eget initiativ. Det framgår att vissa lärare har fått möjligheten att utbilda sig genom sina arbetsplatser, medan andra fått göra det på sin egen fritid. Studien synliggör även lärares uppfattningar om det stöd de fått under införandet av programmering, vilket talar för att majoriteten av lärarna inte varit nöjda. Resultatet visar att lärarnas kunskaper om programmering hindrar dem från att bedriva en undervisning för alla elevers olika kunskapsnivåer.Studien bidrar således till att uppmärksamma lärares egna uppfattningar om deras kunskaper och undervisning om programmering. Rektorer och lärare kan därför, genom denna studie, bedöma om det behöver läggas mer fokus eller resurser på kompetensutveckling av lärare.

The goal with this assignment has been to study the product Spotalike and develop a new version to make the product more attractive. The studying of the product was done with the help of user data, such as how Spotalike is being used, what target audience it has, why it's being used, etc. The new version of Spotalike was planned by making design sketches and prototypes which were created as a first step in order to get a better picture of what the result would be. The new version is not available to the public, but it is fully functional and works locally. The solution that was concluded was to develop a music player which is built on the founding principles of the old Spotalike. The music player is developed with React and is powered by Spotify. Besides the old functions there are also new functions that has been implemented, and the interface has been redesigned. There is currently no new user data available to determine the result of the development, since the new version of Spotalike hasn't been made public yet. / Målet med detta projektarbete har varit att granska produkten Spotalike och utveckla en ny version som gör produkten mer eftertraktad. Problemgranskning har gjorts med analysering av användardata – hur Spotalike används, av vem, varför den används, osv. Den nya versionen av Spotalike planerades med hjälp av designskisser och prototyper som togs fram som första steg för att få en bättre bild av slutresultatet. Den nya versionen är inte tillgänglig publikt, men är fullt funktionell lokalt. Lösningen som togs fram var att skapa en musikspelare som bygger på de grundprinciper gamla Spotalike har. Musikspelaren är bygd med React och använder Spotify i bakgrunden som motor. Utöver de redan befintliga funktionerna så har även nya funktioner tagits fram och gränssnittet har redesignats. Någon ny användarstatistik för att se om lösningen har gynnat bra resultat i form av användarupplevelse finns inte tillgänglig då tjänsten ännu inte har hunnit bli tillgängligt publikt.

The fuel usage of a hybrid electric vehicle can be reduced by strategically combining the usage of the combustion engine with the electric motor. One method to determine an optimal split between the two is to use dynamic programming. However, the amount of computations grows exponentially with the amount of states which makes its usage difficult on sequential hardware. This thesis project explores the usage of FPGAs for speeding up the required computations to possibly allow the optimisation to run in real time in the vehicle. A tool to convert a vehicle model to a hardware description language was developed and evaluated. The current version does not run fast enough to run in real time, but some optimisations which would allow that are proposed.

Digital services involving large systems with multiple users are ubiquitous in modern society. The systems are often complicated and made up of multiple devices and communication protocols. A fundamental problem in this context is how the behavior of a system changes as the number of users vary. In particular, when do the systems’ resources saturate and how does the system behave when close to saturation. Performance testing is key for addressing this fundamental problem. Performance testing is the scope of this project. Performance tests can be used for inference of, for example, a system's scalability. Furthermore, it can be used to provide general guarantees on the services that can be delivered. Performance testing at the company Phoniro AB is considered. The platform Phoniro Care is the back-end service for the company’s products. The Phoiro 6000 system is one of the products that uses Phoniro Care. The system allows for multiple users and offers alarm services. The primary focus of this project is to determine the behavior of that system during varying levels of simulated load, and furthermore analyze the data extracted from such simulations and tests. The open source software JMeter was used as the tool for performance testing. It was selected from a set of candidate tools that have been evaluated in the literature based on various performance criteria. The results are presented by graphs showing the time evolution of different performance indicators. A conclusion from this work is that the implemented performance testing framework helps to answer questions about the systems’ behavior. Questions that are important for the company’s further development and expansion of the system. Furthermore, the proposed framework establishes a foundation for further inquiries on the subject.

Identifying semantic relations in natural language text is an important component of many knowledge extraction systems. This thesis studies the task of hypernym discovery, i.e discovering terms that are related by the hypernymy (is-a) relation. Specifically, this thesis explores how state-of-the-art methods for hypernym discovery perform when applied in specific language domains. In recent times, state-of-the-art methods for hypernym discovery are mostly made up by supervised machine learning models that leverage distributional word representations such as word embeddings. These models require labeled training data in the form of term pairs that are known to be related by hypernymy. Such labeled training data is often not available when working with a specific language domain. This thesis presents experiments with an automatic training data collection algorithm. The algorithm leverages a pre-defined domain-specific vocabulary, and the lexical resource WordNet, to extract training pairs automatically. This thesis contributes by presenting experimental results when attempting to leverage such automatically collected domain-specific training data for the purpose of domain adaptation. Experiments are conducted in two different domains: One domain where there is a large amount of text data, and another domain where there is a much smaller amount of text data. Results show that the automatically collected training data has a positive impact on performance in both domains. The performance boost is most significant in the domain with a large amount of text data, with mean average precision increasing by up to 8 points.

Most research behind the use of Machine Learning models in the field of Human Activity Recognition focuses mainly on the classification of daily human activities and aerobic exercises. In this study, we focus on the use of 1 accelerometer and 2 gyroscope sensors to build a Deep Learning classifier to recognise 5 different strength exercises, as well as a null class. The strength exercises tested in this research are as followed: Bench press, bent row, deadlift, lateral rises and overhead press. The null class contains recordings of daily activities, such as sitting or walking around the house. The model used in this paper consists on the creation of consecutive overlapping fixed length sliding windows for each exercise, which are processed separately and act as the input for a Deep Convolutional Neural Network. In this study we compare different sliding windows lengths and overlap percentages (step sizes) to obtain the optimal window length and overlap percentage combination. Furthermore, we explore the accuracy results between 1D and 2D Convolutional Neural Networks. Cross validation is also used to check the overall accuracy of the classifiers, where the database used in this paper contains 5 exercises performed by 3 different users and a null class. Overall the models were found to perform accurately for window’s with length of 0.5 seconds or greater and provided a solid foundation to move forward in the creation of a more robust fully integrated model that can recognize a wider variety of exercises.

This essay describes the work to document and evaluate tools for front-end development at the company B3 Consulting Group. The methods that are evaluated is AMP-stack and JAM-stack. AMP-stack uses Apache, MySQL and PHP, Wordpress is used to build with that method. JAM-stack uses JavaScript, API and Markup, for that method GatsbyJS and Contentful is used. The evaluation is done both through user tests and how the methods are percived from a developer perspective. The project also includes work with customers at the company the project is done, Wordpress is used for those web pages. / Denna rapport beskriver arbetet med att dokumentera och utvärdera verktyg för frontend-utveckling vid företaget B3 Consulting Group. Metoderna som utvärderas är AMP-stack och JAM-stack. AMP-stack bygger på Apache, MySQL och PHP, för att bygga med den metoden har Wordpress använts. JAMstack bygger på JavaScript, API samt Markup, för den metoden har GatsbyJS samt Contentful använts. Utvärdering sker både genom användartester och hur metoderna upplevs ur ett utvecklarperspektiv. I arbetet ingår även arbete med kunder till företaget projektet utförts hos, för deras webbplatser har Wordpress använts.

Chatbots are computer programs with the capability to lead a conversation with a human user. When a chatbot is unable to match a user’s utterance to any predefined answer, it will use a fallback intent; a generic response that does not contribute to the conversation in any meaningful way. This report aims to investigate if a content-based recommender system could provide support to a chatbot agent in case of these fallback experiences. Content-based recommender systems use content to filter, prioritize and deliver relevant information to users. Their purpose is to search through a large amount of content and predict recommendations based on user requirements. The recommender system developed in this project consists of four components: a web spider, a Bag-of-words model, a graph database, and the GraphQL API. The anticipation was to capture web page articles and rank them with a numeric scoring to figure out which articles that make for the best recommendation concerning given subjects. The chatbot agent could then use these recommended articles to provide the user with value and help instead of a generic response. After the evaluation, it was found that the recommender system in principle fulfilled all requirements, but that the scoring algorithm used could achieve significant improvements in its recommendations if a more advanced algorithm would be implemented. The scoring algorithm used in this project is based on word count, which lacks taking the context of the dialogue between the user and the agent into consideration, among other things.