There are a number of techniques for automatic bug detection, most of them have a high false positive rate when used in practice. This work proposes an approach, named Strip-Miner, that combines simple dependency analysis of code with a data mining technique "frequent itemset mining" to reduce the false positive rate. We adopt a two phase approach 1) finding the potential bugs and 2) filtering the false positive ones. In the first phase we extract code elements and dependencies among them using static analysis and frequent itemset mining to find programming patterns where a deviation from these patterns is considered as a potential bug. In the second phase, we use the extracted dependencies to build dependency chains between program elements in a programming pattern and a lack of such a chain potentially makes a bug false positive. Our evaluation on a set of 7 benchmarks consisting of large software code including OpenSSL, PostgreSQL, Git, FFMPEG, SQLite, Binutils and Putty shows that combining simple de- pendency analysis with pattern mining can significantly decrease the number of generated bugs. Using our approach we are able to reduce the number of generated bugs by up to 99.9% with a false positive rate of 65.19% and true positive rate of 34.18% on average as compared to an earlier frequent itemset mining based approach "PR-Miner". / Master of Science / Every software code has bugs in it that can change its expected behavior. There have been a lot of efforts to automate the process of bug detection but most of the techniques proposed have a high rate of false alarms. Some of these techniques leverage the information available in software code to extract programming patterns that can be used to find potential bugs. Although such an approach has proved to be fruitful for finding bugs but large number of false alarms makes it almost useless in software development. The elements present in a software code have relationships among them formally known as dependencies and the process of finding them is known as dependency analysis. There is a technique known as market basket analysis used by large retailers to find association between items. It works by looking for combinations of items that occur together frequently in transactions. Similarly, in a software code combinations of elements that occur together, can be used to find association between them. This technique is formally known as frequent itemset mining in the data mining domain. This work proposes an approach, named Strip- Miner, that combines dependency analysis with frequent itemset mining to reduce the rate of false alarms. We adopt a two phase approach 1)finding the potential bugs in code and 2)filtering the false alarms. In the first phase we extract code elements and dependencies among them and use frequent itemset mining to find programming patterns where a deviation from these patterns is considered as a potential bug. In the second phase, we use the extracted dependencies to build dependency chains between program elements present in a programming pattern and lack of such a chain is an indication of false alarm. Our evaluation on a set of 7 benchmarks consisting of large software code including version control systems, database management systems, software security libraries and utility software like media players shows that combining simple dependency analysis with frequent itemset mining can significantly decrease the rate of false alarms. Using our approach we are able to reduce the number of generated bugs by up to 99.9% with a false alarms rate of 65.19% and real bugs rate of 34.18% on average as compared to an earlier frequent itemset mining based approach "PR-Miner".

We discuss a new approach to the construction of software systems. Instead of attempting to build a system that is as free of errors as possible, the designer instead identifies key properties that the execution must satisfy to be acceptable to its users. Together, these properties define the acceptability envelope of the system: the region that it must stay within to remain acceptable. The developer then augments the system with a layered set of components, each of which enforces one of the acceptability properties. The potential advantages of this approach include more flexible, resilient systems that recover from errors and behave acceptably across a wide range of operating environments, an appropriately prioritized investment of engineering resources, and the ability to productively incorporate unreliable components into the final software system. / Singapore-MIT Alliance (SMA)

No description available.

This dissertation entitled ’Model-Based Exploration of Parallelism in the Context of Automotive Multi-Core Systems’ deals with the analytical investigation of different temporal relationships for automotive multi-processor systems subject to critical, embedded, real-time, distributed, and heterogeneous domain requirements. Vehicle innovation increasingly demands high-performance platforms in terms of, e.g., highly assisted or autonomous driving such that established software development processes must be examined, revised, and advanced. The goal is not to develop application software itself, but instead to improve the model-based development process, subject to numerous constraints and requirements. Model-based software development is, for example, an established process that allows systems to be analyzed and simulated in an abstracted, standardized, modular, isolated, or integrated manner. The verification of real-time behavior taking into account various constraints and modern architectures, which include graphics and heterogeneous processors as well as dedicated hardware accelerators, is one of many challenges in the real-time and automotive community. The software distribution across hardware entities and the identification of software that can be executed in parallel are crucial in the development process. Since these processes usually optimize one or more properties, they belong to the category of problems that can only be solved in polynomial time using non-deterministic methods and thus make use of (meta) heuristics for being solved. Such (meta) heuristics require sophisticated implementation and configuration, due to the properties to be optimized are usually subject to many different analyses. With the results of this dissertation, various development processes can be adjusted to modern architectures by using new and extended processes that enable future and computationally intensive vehicle applications on the one hand and improve existing processes in terms of efficiency and effectiveness on the other hand. These processes include runnable partitioning, task mapping, data allocation, and timing verification, which are addressed with the help of constraint programming, genetic algorithms, and heuristics.

Im Rahmen der Dissertation wurden die Wechselwirkungen, mögliche Potenziale und Herausforderungen von agilen Prinzipien im Hinblick auf IoT– und KI–Anwendungen auf den Softwareentwicklungsprozess in mittelständischen Unternehmen untersucht. Dazu wurde primär dem gestaltungsorientierten Forschungsparadigma, welches von besonderer Relevanz für die konstruktionsorientierte deutsche Wirtschaftsinformatik ist, gefolgt. Ein Primärziel dieses Forschungsvorhabens war die theoretische Herleitung agiler Methoden und Technologien zur Verbesserung der Software–Engineering–Prozesse mittelständischer Systemhäuser sowie die empirische Validierung und kritische Betrachtung dieser Technologien und Methodiken anhand ausgewählter, repräsentativer Fallbeispiele. Das zweites Primärziel der Dissertation ist die Forschungsarbeit an KI–Systemen, insbesondere im Rahmen von IoT–Anwendungen unter Berücksichtigung der möglichen Wechselwirkungen.