DE / EN

SM 453/IS 722/IS 918 Bachelor and Master Seminar

Course Information
Lecturer Prof. Dr. Christian Becker
Type Seminar (SM 453 for Bachelor & IS 722 for Master & IS 918 for MMBR)
Credit Points 6 ECTS (MMM & MMBR), 5 ECTS (B.Sc. WI since HWS 2013), 4 ECTS (M.Sc. WI since HWS 2013 and B.Sc. WI before HWS 2013)
Prerequisites Basic knowledge in information technology
Course Language English
Form of assessment Conference style seminar (see details below)
Registration Please see information below!
Martin Breitbach, M.Sc.

Martin Breitbach, M.Sc.

Contact person for Seminar

For further information please contact Martin Breitbach.

  • Schedule

    Registration July 15 – September 06
    Confirmation until Wed September 08 (noon)
    De-Registration possible until Thu September 09 (noon)
    Kick-Off Meeting Mon September 13
    First paper draft deadline Fri October 29 (noon)
    Review deadline Fri November 12 (noon)
    Camera ready (final) paper deadline Fri November 19 (noon)
    Conference (final presentations) Fri November 26
  • Conference style seminar

    This seminar is organized in a scientific conference style. All accepted participants must write a scientific paper about the assigned topics and submit those papers until the first draft deadline. After that, the paper review phase starts and each paper will be assigned to at least two other participants who have to review the papers of two or three other authors. After the review phase, the reviews must be submitted to the supervisors, which distribute them to the paper authors. After that, the authors have time to improve their papers based on the feedback from the reviews, before they need to hand in their camera-ready (final) version of the paper. At the end of the semester, the “conference” (final presentations) will take place.

    The grading is divided into different parts: The first part is the camera ready version of the seminar paper. This is the most important part and it is weighted with 50% of the overall grade. Second, the reviews for the other authors  are weighted with 20%. It is crucial to look at the work of others with a critical eye and to give constructive feedback. The last grading criterion is the presentation at the “conference” and the participation during the discussions (30%).

    Attendance at the kick-off session and the final presentation session is mandatory.

    All papers must use the IEEE manuscript template. We offer a customized version of the template here.

  • Registration

    Please apply via our online registration tool only (accessible inside the university network via VPN) . Registration is possible from July 15 until September 06.

    Requirements:

    • CV
    • Transcript of records

    We will not consider registrations via e-mail or incomplete registrations in the registration tool.

     

  • Topics

    01 – “Your Eyes Tell Me That You Are Tired” – A Survey on Gaze-based Cognitive Load Prediction.

    Supervisor: Melanie Heck

    Using eye tracking to evaluate attentional parameters has become an established practice in market research. As the technology becomes more mature, companies like Apple and Google have worked on integrating eye tracking functionalities into their devices. This opens up new opportunities to unobtrusively track the user’s attention and learn something about not only their interests, but also their cognitive state. By predicting cognitive load measures such as workload, fatigue, or mind wandering, adaptive systems can dynamically react to information overload or loss of concentration.

    The objective of this seminar paper is to conduct a structured literature review on gaze-based cognitive load prediction. It should investigate what cognitive load measures can be predicted with eye tracking, and what issues have so far been the focus of cognitive load prediction research.

     

    02 – Making E-Learning More Effective by Adapting to the Student’s Gaze.

    Supervisor: Melanie Heck

    As the ongoing Covid-19 pandemic has pushed e-learning platforms into the spotlight, the systems have not only massively improved the provided content, but have also started to provide personalized learning schedules and methods for each student. Adaptation is mostly based on feedback from the student’s performance on exercises and tests. Much more information about the students’ cognitive states, interests, and personality could be inferred from eye tracking. Such information could then be used to retrieve learning content that fits the student’s general preferences, learning style, and current state of mind.

    The objective of this seminar paper is to conduct a structured literature review on gaze-based e-learning systems. The literature analysis should reveal which issues are currently driving research, and identify gaps in the literature that provide research opportunities.  

     

    03 – “How Do I know How Much Work You Need?” – Using Biosignals to Infer Cognitive Load.

    Supervisor: Melanie Heck

    Adaptive systems support the user by always showing the right content at the right time, and in the right media format. But in order to know what is “right” for the user at a specific moment, the system needs to know how much idle cognitive capacity the user has at the moment. Biosignals such as eye tracking, EEG, or ECG provide insights into a person’s nervous system, and can therefore be used to predict cognitive load.

    The objective of this seminar paper is to analyze survey papers on cognitive load predictions. It should investigate what aspects are considered in each meta-analysis and identify whether the focus of prior research is different for each biosignal.

     

    04 – Proactive Self-Adaptive Systems

    Supervisor: Michael Matthé

    Self-adaptive systems have the ability to modify their configuration in response to changes in their environment. Intelligent transportation systems (e.g. platooning, vehicle networks) and cloud computing are examples of application areas where self-adaptive systems may be employed. These systems are context-aware allowing them to monitor their environment (context) and adapt their system configuration to improve performance. Two approaches exist to perform the adaptation of the system configuration – reactive and proactive. A “traditional” reactive system monitors its context and performs an adaptation of its configuration after detecting the need for it. Contrarily, a proactive system aims to predict a future change in context and preemptively reconfigure to a more optimal system configuration.

    The goal of this seminar paper is to provide an overview of use cases and application areas in which proactive self-adaptive systems are employed. Different methods for predicting context changes, such as reinforcement learning or Markov chains, should be compared.

     

    05 – Decentralized Adaptation in Self-Adaptive Systems

    Supervisor: Michael Matthé

    Traditionally Self-Adaptive Systems (SAS) consist of a managed resource and an adaptation logic comprised of a MAPE-K loop (Monitor, Analyze, Plan, Execute – Knowledge). SASs are commonly made up by a system of systems. For example, in a platooning use case, a single platoon may consist of one controlling system (e.g. the lead vehicle) and multiple following vehicles. The adaptation logic of each of the vehicles following the controlling vehicle may be reduced to just the monitoring and execution component, moving the rest of the components to the lead vehicle. In a centralized adaptation approach, one adaptation logic is used to manage all system resources. Contrarily, in a fully decentralized adaptation approach, each managed resource has its own adaptation logic. Additionally, hybrid approaches may employ a combination of centralized and decentralized adaptation approaches or switch between the approaches dynamically at runtime.

    The objective of this seminar paper is to perform a literature overview of decentralized or hybrid adaptation approaches in self-adaptive systems. The advantages and challenges of decentralized and hybrid approaches in comparison to the centralized approach should be outlined.

     

    06 – Context Inference in Context-Aware Systems

    Supervisor: Michael Matthé

    Context-aware systems can monitor their environmental context and perform adaptations of system configuration depending on the detected context. An example of an area of applications that may benefit from context-awareness are mobile devices. Using sensor data and network information these devices may detect context information such as their location and time and optimize system behavior to improve performance and user satisfaction. Due to the dynamic nature of the environment of such systems, the context of the system may change randomly or during periodical intervals. Context inference allows a system to make predictions or assumptions regarding these future changes in its context, thereby enabling it to more proactively adapt.

    The aim of this seminar paper is to perform a literature review of context-inference for context-aware adaptive systems. This should include an overview of different methods for both performing the collection of context data as well as methods for performing the analysis and inference of the collected data (e.g. reinforcement learning).

     

    07 – Task Allocation in Distributed Systems

    Supervisor: Melanie Brinkschulte

    For the efficient use of system resources in a distributed system, task allocation including placement and migration play an important role. Besides the beneficial effect of an even load balancing for the performance of the overall system also the communication overhead can be reduced by locating tasks with intensive communication among them on the same host. In Addition, fault tolerance can be achieved in case of failure or overload of a host or even a communication channel.

    The objective of this seminar paper is to analyze existing task allocation, placement and migration approaches in distributed systems. Thereby, three questions should be investigated: (1) How and based on which criteria can the best fitting host for a task be determined? (2) When should a task be migrated? (3) How can the overall system decide which host takes over a task which need to be migrated?

     

    08 – Evaluation of Self-Adaptation

    Supervisor: Melanie Brinkschulte

    The complexity of distributed systems is increasing in recent years, leading to greater effort for design, operation, optimization and maintenance of such systems. As a result, self-adaptation is an increasingly important capability for many systems to achieve fault tolerance and optimization. Thereby, the system needs a metric in order to decide how to adapt themselves: e.g. rules. 

    The objective of this seminar paper is to analyze how different self-adaptation strategies are evaluated in literature. Thereby three questions should be investigated. (1) How is self-adaptation realized? (2) Which evaluation tools exist? (3) Which criteria can be used to evaluate the effectiveness of the self-adaptation within the system?

     

    09 – User Interaction to Support Environmental Behaviour

    Supervisor: Sonja Klingert

    The energy turnaround is turning the formally centralized energy system into a distributed energy system. This poses a high level of challenges to the orchestration of distributed electricity generation. This transformation will not be possible without a behaviour change from the part of the users of electricity. But how to make people aware of current problems? There are different modes of interaction and the seminar paper should give an overview on the current status.