Pro-/Haupt- und Forschungsseminar VSR (SS 2024)

Questions:

• What are GOMS/KLM Models? How do they work? Why are they used? What is the (data) basis on which they were created?
• For what kinds of user interfaces can they be or have they been applied? What are their limitations?
• Apply GOMS modeling to real world examples (e.g. vsr website) and demonstrate how they can be used to improve these interfaces.

Literature:

• https://cogulator.io/
• https://syntagm.co.uk/design/klmcalc.shtml
• https://www.cogtool.org/
• Card, S. K., Moran, T. P., & Newell, A. (1983). The psychology of human-computer interaction. Hillsdale, N.J. : L. Erlbaum Associates.
• Kieras, D. (1997). A guide to GOMS model usability evaluation using NGOMSL (Chapter 31). In M. Helander, T.K. Landauer & P.V. Prabhu (Eds.), Handbook of Human-Computer Interaction. Amsterdam: North-Holland Elsevier Science Publishers. Kim,
• John, B. and Kieras, D. The GOMS family of user interface analysis techniques: comparison and contrast. ACM TOCHI, 3 (4). 1996. 320-351.
• John, B. E. (2010). CogTool: Predictive human performance modeling by demonstration. 19th Annual Conference on Behavior Representation in Modeling and Simulation 2010, BRiMS 2010, 308–309.

Questions:

• What is signposting?
• How can signposting be applied to improve machine-readbility of resources on the Web?

Literature:

• https://signposting.org/
• https://signposting.org/FAIR/
• Own research

Questions:

• What are typical characteristics of HTTP versions?
• TCP, UDP, QUIC? What is used how below HTTP and how does it influence the protocol and its usage?
• A technical deep dive on demonstrators for HTTP version comparison with focus on HTTP/3 is expected.

Literature:

• own research

Questions:

• What are RDFSurfaces? Give a technical deep dive.
• How can RDFSurfaces help to improve decentralized knowledge graphs?

Literature:

• https://w3c-cg.github.io/rdfsurfaces/
• https://arxiv.org/abs/2305.08476

Questions:

• Create a corpus of Customer Service Chatbots that you analyze on certain criteria. The corpus should include at least 20 different chatbots (40 for Hauptseminar). For this you will get access to a gitlab repository in which you can save your results. Build a strategy on how to systematically search for these customer service chatbots and describe this methodology.
• The corpus should include at least information about the corporation, it's field, the source, how the chatbot introduces itself, if available information on the trained data, possible access restrictions, information on the user interface, etc.
• Additionally have a look at the code and categorize it. The following questions can be used as an idea for criteria and added on - use the available literature to formulate specific criteria: Is the code openly accessible and readable? Are third party chatbots used? If yes, which ones are used? Are the chatbots included via API-call etc. from the outside? How are they called on in the code?

Literature:

• own research
• Adamopoulou, E., Moussiades, L. (2020). An Overview of Chatbot Technology. In: Maglogiannis, I., Iliadis, L., Pimenidis, E. (eds) Artificial Intelligence Applications and Innovations. AIAI 2020. IFIP Advances in Information and Communication Technology, vol 584. Springer, Cham. https://doi.org/10.1007/978-3-030-49186-4_31
• Adamopoulou, Eleni, and Lefteris Moussiades. "Chatbots: History, technology, and applications." Machine Learning with Applications 2 (2020): 100006. https://doi.org/10.1016/j.mlwa.2020.100006
• Akma, N., Hafiz, M., Zainal, A., Fairuz, M., Adnan, Z.: Review of chatbots design techniques. Int. J. Comput. Appl. 181, 7–10 (2018). https://doi.org/10.5120/ijca2018917606
• M. Baez, F. Daniel, F. Casati and B. Benatallah, "Chatbot Integration in Few Patterns," in IEEE Internet Computing, vol. 25, no. 3, pp. 52-59, 1 May-June 2021, doi: 10.1109/MIC.2020.3024605.

Questions:

• Intro: As a data engineer and web-developer it is inevitable to understand the concepts of data privacy and licensing and its significance in today’s data- and software-driven world.
• Introduction and basic definitions: data, processing,
• Personal Data: definition, types of personal data, processing principles, anonymization and pseudo-anonymization, mixed data, documentation, data subjects’ rights
• Introduction and basic definitions: licenses
• Open Sources Licenses: definition, types of Open Source Licenses, overview of the most popular examples of Open Source Licenses with the respective conditions, compatibility of licenses (for instance in context of data fusion), review available websites and tools for Open Source License Management, including demonstrations
• Optional: - Discuss some legal challenges that arise for training data usage in connection with AI-models
• Optional: - Name and shortly explain current or planned legislation for AI applications

Literature:

Questions:

• How does Bun compare to its competitors? What does Bun do differently in terms of functionality?
• How does the architecture of Bun differ from Node.js and Deno?
• Compare the three runtimes - Deno, Bun and Node.js - by running different types of Benchmarks. Measure performance and memory consumption!

Literature:

• Own research

Questions:

• • Give a short introduction of the explainable AI (XAI) and Knowledge Graphs (KGs) and an short introduction of the role of KGs in XAI.
• • State and compare the respective opportunities and challenges of KGs in XAI.
• • Where is potential for an integration of both approaches and what would be the benefit of such synergies for various applications? (For ex. What is the benefits of using KGs in explainable recommendation systems)
• • Explain some ways KGs could be or already have been used to improve on XAI.
• • Give a more detailed review, supplemented by a short demonstration of one successful realization of one combined approach.
• • What are future trends in the XAI field?

Literature:

• 1. Lecue, F. (2020). On the role of knowledge graphs in explainable AI. Semantic Web, 11(1), 41-51.
• 2. Rajabi, E., & Etminani, K. (2022). Knowledge-graph-based explainable AI: A systematic review. Journal of Information Science, 01655515221112844.
• 3. Tiddi, I., & Schlobach, S. (2022). Knowledge graphs as tools for explainable machine learning: A survey. Artificial Intelligence, 302, 103627.

Questions:

• What are common methodologies for anti-spam?
• What are the recent years' changes on anti-spam?
• How does anti-spam tackle generative AI?

Literature:

• own research

Questions:

• What is the current State of the Art in constructing Knowledge Graphs from unstructured text?
• How can meaningful and appropriate entities be extracted with LLMs? Which Prompting Techniques help in improving the output of the model wrt. Entity Extraction?
• LLMs fall under the category of Generative AI. Hence, is it feasible and appropriate to employ LLMs to determine Entities that are not in the text, but can be logically derived from it?
• How do the approaches compare with standard NLP tools?

Literature:

• Own research