Topics for Theses
Here you will find a selection of topics for bachelor’s and master’s theses that we offer at the Chair of Human-Centered Information Systems. These topics are intended primarily as suggestions and can be adapted as needed based on your interests and academic focus. The specific research question and scope will be determined jointly through discussion, depending on the type of thesis (Bachelor’s or Master’s).
If you are interested in one of the topics, please feel free to contact us through Prof. Dr. Thomas Ludwig or the respective thesis advisor.
Topics for Bachelor's and Master's Theses (Business Information Systems, Human-Computer Interaction, and Computer Science)
For dynamically responsive virtual characters, text-to-speech serves as a central interface between a response generated by an LLM and the participants’ actual perception of it. Particularly in interactive simulations, it is not only speech quality that is crucial. Latency, streaming capability, local executability, required hardware, multilingual support, and the ability to convey emotional nuances through speech can also play an important role.
As part of this thesis, various current text-to-speech systems will be systematically compared and evaluated using a suitable test environment. In addition to objective technical metrics, subjective criteria such as naturalness and intelligibility will also be considered.
Broad research question:
Which text-to-speech methods are suitable for dynamic LLM-based characters in real-time simulations in terms of latency, speech quality, resource requirements, and integrability?
Key Focus Areas:
- Text-to-Speech
- Large Language Models
- Speech AI
- Benchmarking
- Unity / Unreal Engine
Advisor: Alexander Thier
In VR training scenarios, virtual characters should not merely execute pre-scripted sequences, but should be able to react dynamically to situations and participants. Large language models can generate dialogues, action intentions, or structured states for an NPC. However, linguistic output alone is not sufficient for a credible simulation: statements and decisions must be accompanied by appropriate body postures, gestures, gaze directions, and animations.
The goal of this work is to develop and investigate an architecture that can translate structured LLM outputs into appropriate avatar animations. In particular, the study will examine what information an LLM must provide, how this information can be reliably interpreted, and how animations can be synchronized with speech and behavior.
Broad research question:
How can the output of a large language model be automatically converted into believable and context-dependent avatar animations?
Key Areas
- Unity / Unreal Engine
- Large Language Models
- Avatar Animation
- AI
- NPC Systems
Advisor: Alexander Thier
Complex training simulations require a control center from which an operator can monitor and influence the ongoing scenario. With up to 16 participants taking part simultaneously, this gives rise to numerous possible functions: starting and ending scenarios, monitoring participants, triggering events, changing environmental conditions, controlling virtual characters, performing technical diagnostics, or flagging relevant situations for later debriefing.
This project will first examine what information and control options an operator actually needs and how these can be structured effectively. Building on this, a prototype UI and control concept will be developed and, if appropriate, integrated directly into the existing simulation platform. In addition to the implementation itself, the focus is particularly on information architecture, usability, and the avoidance of unnecessary complexity.
Broad research question:
What information and control options does an operator need to carry out complex multi-user VR training scenarios, and how can these be presented in a suitable interface?
Key Areas
- UI/UX
- Operator Interface
- Unity / Web UI
- Multi-User VR
- Requirements Engineering
Advisor: Alexander Thier
Manually creating new training environments is time-consuming and limits the number of different scenarios. A procedural content generation system could instead automatically or semi-automatically generate different operational environments from reusable building blocks. Possible examples include residential buildings, office spaces, warehouses, road sections, or other operational sites, whose floor plans, rooms, objects, and hazard sources are generated based on configurable rules.
The initial goal of this work is to develop the basic architecture of such a system. The focus is less on the complete generation of highly complex operational scenarios and more on the question of how environments can be described modularly and generated reproducibly according to defined rules. Building on this, parameters such as room structure, doors, furnishings, operational objects, or hazard areas can be varied.
Broad research question:
How can a modular procedural content generation system be designed that automatically generates diverse, plausible operational environments suitable for VR training?
Key Areas
- Procedural Content Generation
- Unity / Unreal Engine
- Level Generation
- Modular Assets
- Rule-Based Systems
Advisor: Alexander Thier
Abstract
This thesis examines how an in-house language model can be developed and integrated as a central platform for technical and organizational expertise at GIZEH Verpackungen GmbH & Co. KG. The focus is on whether existing process data, quality guidelines, tool documentation, and machine settings can be harnessed through approaches such as retrieval-augmented generation or targeted fine-tuning in a way that provides employees with fast and reliable access to relevant experiential knowledge in their day-to-day production work. To this end, the technical prerequisites, data protection requirements, and typical sources of error—such as hallucinations—are analyzed and tested in a prototype or feasibility concept.
Research Question
What technical, data-related, and organizational conditions determine the feasibility of an internal large language model as a knowledge base in a plastics-processing industrial company?
Possible sub-questions include:
What system architecture allows for the reliable and source-accurate integration of heterogeneous process and experiential knowledge?
How must the interaction design be structured so that employees in production, toolmaking, and quality management accept and use the system in their daily work?
What requirements regarding data security, operating costs, and validation routines must be met to minimize erroneous responses in a manufacturing context?
As part of this thesis, the existing information flow at GIZEH Verpackungen will be mapped, an architectural concept will be developed, and a prototype will be tested using selected use cases. The goal is to provide a well-founded decision-making framework as well as concrete recommendations for the step-by-step implementation of the system.
Recommended Prerequisites
Basic understanding of language models and RAG architectures
Interest in knowledge management, information retrieval, or data modeling
Interest in industrial production processes, ideally in the field of injection molding or deep drawing
Willingness to conduct practical requirements analysis in collaboration with business departments
Abstract
This study examines whether multiple LLM-based agents with different roles can better encourage visitors to reflect on their cultural experience after a museum visit than a single chatbot.
To this end, a prototype will be developed in which, for example, a “curator” provides background information, a “critic” encourages critical engagement, and an “artist” offers creative perspectives. This prototype will be compared with a comparable single-agent chatbot.
Research Question
How does the reflection on a museum experience differ when interacting with a single LLM-based chatbot versus a multi-agent system with different communicative roles?
Possible sub-questions include:
- How do users perceive the different agent roles?
- What differences emerge in the nature and depth of reflection?
- Which design features of a multi-agent system support users’ engagement with their own museum experience?
Both variants will be compared in a user study. The interactions can be analyzed, for example, based on chat logs and subsequent interviews.
The goal is to derive design recommendations for LLM-based multi-agent systems for the digital mediation of cultural content.
Recommended Prerequisites
- Fundamentals of HCI and user research
- Chatbot and conversational UI design
- Initial experience with LLMs and prompt engineering
- Interest in digital cultural and educational formats
Advisor: Philip Weber
Abstract
This study examines how digital collection systems can be designed to help visitors capture personal memories of a museum or cultural visit and revisit them later.
To this end, an interactive prototype will be developed in which users can, for example, compile images, notes, digital postcards, or information about exhibits they have viewed into a personal collection. The goal is to combine institutional content with users’ own contributions.
Research Question
How does the design of a personalized digital collection system influence the perception and memory of a cultural visit?
Possible sub-questions include:
What types of digital mementos do users find particularly personal and meaningful?
How should a personal collection be designed so that users are eager to revisit it after their visit?
How does the degree of personalization influence the perceived connection to the visit experience?
The developed prototype will be evaluated as part of a user study. In this study, different forms of personalization or ways of presenting personal collections (including AR/VR) can be compared with one another.
The goal is to develop design recommendations for digital collection systems that extend the cultural visitor experience beyond the actual visit.
Recommended Prerequisites
- Fundamentals of HCI and user research
- Interest in UX and interaction design
- Basic experience with prototyping
- Interest in museums, cultural education, or digital memory media
Advisor: Philip Weber
This study examines how the introduction of a chatbot changes day-to-day collaboration in a travel expense department. Rather than focusing exclusively on usability or user satisfaction, the study examines changes in task distribution, coordination, communication, accountability, and handoffs between employees and the chatbot.
To this end, employees in a travel expense department are studied through guided interviews and/or participatory workshops. The study focuses on how work was organized before and after the chatbot’s introduction, which tasks are delegated to the chatbot, in which situations employees must still intervene, and whether new forms of work are emerging. This may include, for example, verifying responses, correcting information, handling exceptions, or coordinating escalations.
The analysis can draw on CSCW concepts such as articulation work, division of labor, coordination mechanisms, invisible work, and collaboration between humans and AI.
Possible research question
How does conversational AI change collaboration when it is embedded in an existing organizational workflow?
Possible sub-questions include:
- How are tasks and responsibilities distributed between employees and the chatbot?
- How does the chatbot influence coordination and communication?
- What new forms of articulation work, invisible work, or corrective work emerge?
- How are incomplete or incorrect responses from the chatbot handled?
- How do handoffs between people and the chatbot affect workflows and responsibilities?
- How do employees adapt the chatbot to their everyday work practices?
The goal of this work is to understand how conversational AI is changing collaboration and to derive design recommendations for better coordination and collaboration between humans and AI.
Recommended Prerequisites
- Basic knowledge of HCI and qualitative user research
- Interest in CSCW and workplace studies
- Experience with interviews, workshops, or thematic analysis
- Basic understanding of chatbots and conversational AI
- Interest in collaboration between humans and AI, as well as in organizational work processes
Advisors: Faisal Mahmood, Philip Weber
Description:
Policy measures such as funding programs, regulatory requirements, or institutional governance structures can initiate, promote, or require collaboration between organizations. Such forms of policy-induced collaboration differ in some respects from voluntarily formed collaborations, as the collaboration may be shaped by external framework conditions, predetermined roles and goals, and institutional requirements.
The specific framework conditions of such collaborations also place demands on the design of digital systems intended to support collaboration. Existing collaboration and CSCW systems have so far taken only limited account of the fact that users do not necessarily collaborate of their own accord or with entirely shared interests.
The aim of this thesis is therefore to examine, from an HCI/CSCW perspective, what specific requirements arise for digital systems designed to support policy-induced collaboration. To this end, the existing state of research on policy-induced or institutionally driven cooperation, as well as on digital support for interorganizational collaboration, will first be reviewed.
Building on this, key challenges and requirements will be identified and structured in the form of design principles or a conceptual framework. Depending on the focus, an interactive prototype may be developed based on this framework to exemplify the implementation of selected design principles, which will then be evaluated.
The thesis aims to establish a connection between research on politically influenced forms of cooperation and the design of digital socio-technical systems.
Research Question:
What requirements arise for the design of digital systems to support policy-induced collaboration?
Objectives:
- To review the existing state of research on policy-induced collaboration and related forms of institutionally shaped interorganizational collaboration,
- Identify key challenges and distinctive features of these forms of collaboration and distinguish them from voluntary interorganizational collaboration,
- Derive requirements for digital collaboration systems,
- To synthesize these requirements into overarching design principles or a conceptual framework,
- Develop and evaluate a conceptual prototype.
Advisor: Robert Fischbach
Digital technologies such as augmented reality (AR), virtual reality (VR), and building information modeling (BIM) are increasingly transforming work processes in the skilled trades. At the same time, user-centered design of such applications is crucial to their acceptance and effectiveness in practice. As part of this project, students analyze existing evaluations of five click prototypes that address different areas of application in the skilled trades and were developed in a university setting.
The prototypes include an AR-supported site inspection, VR-supported orientation sessions, a chatbot for interacting with BIM models, a BIM-based quotation process, and the scanning of environments to compare the actual construction status with the construction plans. The analysis is based on evaluations involving approximately six participants each.
The aim of this thesis is to systematically evaluate the existing evaluation results and compare them with current scientific findings. Building on this, the results will be critically discussed, and concrete proposals for improvement will be developed. Furthermore, new design approaches for the prototypes will be developed, (selectively) implemented, and evaluated. The thesis thus combines empirical analysis, literature review, conceptual design, and implementation.
Advisor: Diana Grüger
Industrial training and working as a skilled professional present a wide range of challenges. Technologies such as augmented reality (AR) and virtual reality (VR) open up new possibilities for supporting and visualizing learning processes and work steps. At the same time, it is crucial to use these technologies effectively so as not to overwhelm skilled workers but rather to provide them with targeted support.
As part of this project, students analyze existing workshop findings from a workshop with industrial trainees on AR/VR applications. The goal is to systematically evaluate needs, usage ideas, and challenges from the trainees’ perspective and compare them with the current state of research. A particular focus is on Diminished Reality (DR), in which real-world environments are selectively reduced or simplified to direct attention to relevant content and support learning processes.
Based on the analysis, students develop their own design concepts for AR/VR applications incorporating DR elements in an industrial context. These concepts should be practical, user-centered, and didactically sound. Subsequently, a prototype that brings the developed concepts to life is created and evaluated.
Advisor: Diana Grüger
Presumably, everyone has to give a presentation to an audience at some point in their lives—whether it’s a presentation in school, a lecture during college or vocational training, or a report to coworkers at work. Yet a significant number of people feel anxious and worried about giving a presentation. VR now makes it possible to practice presentations in an immersive, stress-free, and realistic environment to reduce anxiety.
Building on the findings of an existing bachelor’s thesis, students are further developing and evaluating an AI-supported VR training environment for presentations. The training environment is tailored to an exam situation in front of a panel, and the AI used in the current prototype evaluates various factors related to the presenter.
The goal of this thesis is therefore the critical refinement of an existing prototype based on the student’s own concepts, literature, and/or empirical data, as well as the evaluation and iterative improvement of the prototype. The thesis thus combines empirical analysis, literature review, conceptual design, and implementation.
Advisor: Diana Grüger
Brief Description
The goal of this master’s thesis is to develop a prototype or clickable mockup for a digital knowledge marketplace that supports the exchange of tacit knowledge among trainees. The focus is not on the transfer of documented technical knowledge, but rather on action-oriented experiential knowledge, which is often not recorded in textbooks, work instructions, or other formal documentation.
The concept is based on a “knowledge-for-knowledge” exchange principle. Trainees should have the opportunity to contribute their own experiences, tips, and best practices and, in return, benefit from the knowledge of others. Principles of the Lotus system are used to structure and link the content, thereby making connections between different areas of knowledge visible.
As part of this thesis, the requirements for such a system are analyzed, a user-centered concept is developed, and it is implemented in the form of a clickable prototype. Finally, the extent to which this approach can promote knowledge transfer and knowledge retention within the training program is examined.
Research Question
How can a digital knowledge marketplace be designed to promote the transfer of tacit experiential knowledge among trainees through a reciprocal exchange of knowledge?
Advisor: Christopher Förster
Description:
Abstract AI/ML concepts (e.g., learning, training, bias, or even LLM/NN...) are often difficult for people without a technical background to grasp. This project aims to investigate how such a concept can be made tangible through a physical, interactive prototype. In a participatory process, the target group will help decide which concept to implement; subsequently, a physical prototype will be developed that conveys this concept interactively (literally making it tangible). A comparative user study will examine how learning outcomes and understanding differ when using the prototype compared to traditional teaching methods (e.g., video, text, digital simulation).
Possible research questions:
- Which AI/ML concept is particularly suitable for physical instruction, and why? How can this selection be made in a participatory manner with the target group?
- How must a physical prototype be designed to convey the selected concept in a clear and understandable way?
- How does understanding of the concept differ after using the physical prototype compared to other forms of instruction?
- What role do motivation and engagement play in interactive/haptic education compared to digital/text-based education?
Advisors: Marisa Schmidt, Enzo Frenker-Hackfort
Description:
AI literacy is increasingly becoming a basic requirement for employees in SMEs, yet existing training programs are often too theoretical, too technical, or not tailored to the target audience. This study examines what content and teaching methods are needed to impart basic AI knowledge to employees without a technical background in a practical way. Based on participatory design workshops with the target group, an interactive digital learning hub will be designed and subsequently implemented, for example, as a web prototype. The prototype will be evaluated in a user study, with a particular focus on whether and how participants’ AI literacy improves through its use.
Possible research questions:
- What content and delivery methods do SME employees without a technical background need to build a basic understanding of AI?
- How should an interactive learning hub be designed to convey this content in a way that is tailored to the target audience?
- What impact does participatory design have on the hub’s acceptance and perceived relevance?
- To what extent does using the developed hub measurably improve users’ AI literacy?
Advisors: Marisa Schmidt, Enzo Frenker-Hackfort
Description:
In an inter-company training workshop, the focus is on the transfer of tacit knowledge. That is, knowledge that arises from practical experience and practice and is difficult to articulate explicitly. The goal of this thesis is to examine how gamification can be used to boost trainees’ motivation regarding the digital documentation and sharing of this knowledge among themselves.
As part of this thesis, a systematic literature review will be conducted on the effects of gamification in the context of peer learning, taking into account the specific learning environment of an inter-company training workshop. Scientific studies will be analyzed to identify key findings regarding the promotion of motivation through gamification, proven gamification elements, as well as current trends and challenges. Based on the findings, an initial prototype application will be developed and subsequently evaluated with trainees.
The research questions are:
- Which gamification elements are suitable?
- What potential do these elements offer for promoting motivation with regard to digital knowledge documentation (e.g., using smart glasses) and knowledge exchange among trainees in the inter-company training workshop?
- What are the limitations, and what challenges exist?
Advisor: Stefanie Bächler
Can't find a suitable topic on the list? We always welcome your own topic ideas related to our research areas (e.g., human-AI interaction, socio-technical systems, digital transformation)! Simply contact Prof. Dr. Thomas Ludwig or the relevant team member.
Advisor: Prof. Dr. Thomas Ludwig