Abschlussarbeiten

This master’s thesis aims to conduct a systematic literature review (SLR) on the ethical, legal, and social implications of using AI systems to detect cyberbullying and hate speech, especially in sensitive and security-critical contexts. The focus lies on identifying risks of algorithmic bias, unintended misuse, and discriminatory outcomes, particularly for vulnerable or marginalized social groups.The review will critically assess existing research on data annotation practices, guidelines for responsible AI, and current technical approaches to cyber abuse detection. It will synthesize findings to evaluate how current AI systems align with principles of fairness, accountability, and transparency, and where significant gaps remain.

Literatur:

• Lahby, M., Pathan, A.-S. K., Maleh, Y., & Yafooz, W. M. S. (Eds.). (2022). Studies in Computational Intelligence. Combating Fake News with Computational Intelligence Techniques. Springer International Publishing.
• Nasery, M., Turel, O., & Yuan, Y. (2023). Combating fake news on social media: a framework, review, and future opportunities. Communications of the Association for Information Systems, 53(1), 833-876.
• Kießling, S., Figl, K., & Remus, U. (2021). Human experts or artificial intelligence? Algorithm aversion in fake news detection.

Level: Bachelor

Kontakt: jana.lekscha@uni-bamberg.de

GenAI gained public awareness with the publication of the ChatGPT prototype in late 2022, collecting more than one million users in just five days. There have been many new developments regarding GenAI in the last few years, like e.g., advances in deep learning technology, which have greatly improved the performance or models such as GPT, which have enabled more natural language generation and even improved the ability to generate images. Due to these strong developments the impact GenAI has on us humans increase. While the use of GenAI is already having a particularly strong impact on the private sector, there is still a lot of potential in the world of work that has not yet been fully exploited. In particular, GenAI offers organizations advantages for automating their tasks, being supported in decision-making, creating content, or delivering personalized content.

This thesis focuses on the integration and use of GenAI on work processes in the context of e-commerce. Qualitative research methods should be used to address this question. Thus, expert interviews with employees from an e-commerce department of a company should be conducted. This raises questions on how GenAI is currently being used in a company, what the reasons for this are and what changes the use of GenAI will bring for their employees.

Literatur:
Feuerriegel, S., Hartmann, J., Janiesch, C., & Zschech, P. (2024). Generative ai. Business & Information Systems Engineering, 66(1), 111-126.
Schmidt, C. V. H., Guffler, M., Kindermann, B., & Flatten, T. (2023). Collaborating with Generative AI: Exploring Algorithm Appreciation in Creative Writing.
Tao, Y., Yoo, C., & Animesh, A. (2023). AI Plus Other Technologies? The Impact of ChatGPT and Creativity Support Systems on Individual Creativity.

Level: Bachelor

Kontakt: marie.langer.aic(at)uni-bamberg.de

Generative artificial intelligence (GenAI) in the form of, for example, ChatGPT, Google Bard, or Bing Chat has a significant impact on the design, operation, and application of information systems. This is because generative AI builds on sophisticated natural language processing (NLP) techniques to communicate fluently with humans.Whereas generative AI provides entirely new techniques, for example, to assist the dissemination of information through facilitating information research and text generation, it also poses significant threats, for example, easily and quickly generating persuasive misinformation. The scalability of generative AI could eventually create an unprecedented level of public confusion. Despite that, research has yet to unpack the differences between generative AI misinformation and human-generated misinformation and whether existing solutions to detect misinformation also apply to generative AI.

This thesis aims to provide an overview of AI-generated misinformation. Therefore, a systematic literature review should be conducted to identify the current state of research on generative AI and, specifically, AI-generated misinformation. Further, one joint systematic literature review (SLR) (focussing on comprehensiveness) and/or one critical review (focussing on theory) can be developed. For a master’s thesis, the collection of social media data regarding the extent of and sentiment about AI-generated misinformation in social media and the users’ perception of it can be the next step to identify the status quo in this topic (i.e., SLR/critical review & SM Analysis).

Literatur:
Hofeditz, L., Mirbabaie, M., Holstein, J., & Stieglitz, S. (2021). Do you Trust an AI-Journalist? A Credibility Analysis of News Content with AI-Authorship, ECIS 2021 Research Papers, 50.
Teubner, T., Flath, C. M., Weinhardt, C., van der Aalst, W., & Hinz, O. (2023). Welcome to the era of chatgpt et al. the prospects of large language models. Business & Information Systems Engineering, 65(2), 95-101.
Zhou, J., Zhang, Y., Luo, Q., Parker, A. G., & De Choudhury, M. (2023). Synthetic Lies: Understanding AI-Generated Misinformation and Evaluating Algorithmic and Human Solutions. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems.

Level: Bachelor/Master

Kontakt:marie.langer.aic(at)uni-bamberg.de

Social media platforms have become important channels for disseminating information in times of crisis. Users are looking for specific guidance and real-time information to alleviate feelings of vulnerability. However, the landscape continues to evolve with the increasing presence of social bots, particularly those powered by generative artificial intelligence (GenAI), adding a new facet to crisis communications. While social media is invaluable for urgent interactions, GenAI's inherent tendency to produce inaccurate results poses a challenge for its use in tasks that require precision. In tasks where accuracy is critical, human oversight is crucial, suggesting that augmentation may be a more appropriate strategy than full automation. This research addresses the identification of specific tasks within the functions of GenAI-driven social bots in crisis communication that require human supervision to strike the delicate balance between automation and augmentation.

Literatur:

• Austin, L., Fisher Liu, B., and Jin, Y. 2012. “How Audiences Seek Out Crisis Information: Exploring the Social-Mediated Crisis Communication Model,” Journal of Applied Communication Research (40:2), pp. 188–207. (https://doi.org/10.1080/00909882.2012.654498).
• Bender, E. M., Gebru, T., McMillan-Major, A., and Shmitchell, S. 2021. “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?,” in Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, pp. 610–623. (https://doi.org/10.1145/3442188.3445922).
• Brachten, F., Mirbabaie, M., Stieglitz, S., Berger, O., Bludau, S., and Schrickel, K. 2018. “Threat or Opportunity? - Examining Social Bots in Social Media Crisis Communication,” in Proceedings of the Australasian Conference on Information Systems.
• Maniou, T. A., and Veglis, A. 2020. “Employing a Chatbot for News Dissemination during Crisis: Design, Implementation and Evaluation,” Future Internet (12:12). (https://doi.org/10.3390/FI12070109).
• Ross, B., Pilz, L., Cabrera, B., Brachten, F., Neubaum, G., and Stieglitz, S. 2019. “Are Social Bots a Real Threat? An Agent-Based Model of the Spiral of Silence to Analyse the Impact of Manipulative Actors in Social Networks,” European Journal of Information Systems (28:4), pp. 394–412.
• Ross, B., Potthoff, T., Majchrzak, T. A., Chakraborty, N. R., Ben Lazreg, M., and Stieglitz, S. 2018. The Diffusion of Crisis-Related Communication on Social Media: An Empirical Analysis of Facebook Reactions. (https://doi.org/10.24251/HICSS.2018.319).
• Stieglitz, S., Hofeditz, L., Brünker, F., Ehnis, C., Mirbabaie, M., and Ross, B. 2022. “Design Principles for Conversational Agents to Support Emergency Management Agencies,” International Journal of Information Management (63), (https://doi.org/10.1016/J.IJINFOMGT.2021.102469). Pergamon, p. 102469.

Level: 

• Master: Mixed-Methods-Design - Qualitative analyses (e.g. interviews) and content analysis

Kontakt: jana.lekscha(at)uni-bamberg.de 

The rise of immersive technologies has transformed sports and fitness by offering virtual alternatives to traditional settings (Chen & Li, 2023; Todorov et al., 2019). These developments raise important questions about how such environments influence user motivation. Self-Determination Theory (SDT) provides a robust framework to examine the types of motivation (amotivation, extrinsic, and intrinsic) that shape sustained engagement and performance in these settings (Ryan & Deci, 2017). This thesis aims to explore how Metaverse sports activities impact user motivation and fitness engagement. It aims to elaborate how virtual environments foster different forms of motivation and their subsequent effects on well-being, performance, and long-term training sustainability.

Literature: 

• Chen, H., & Li, H. (2023). Emotional Experience in Virtual Reality Sports Use.
• Deci, E. L., & Ryan, R. M. (2012). Self-Determination Theory. In Handbook of Theories of Social Psychology: Volume 1. SAGE Publications Ltd. https://doi.org/10.4135/9781446249215
• Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in motivation, development, and wellness. Guilford publications.
• Todorov, K., Manolova, A., & Chervendinev, G. (2019). Immersion in Virtual Reality Video Games for Improving Physical Performance Measures: A Review. 2019 27th National Conference with International Participation (℡ECOM), 35–38. https://doi.org/10.1109/℡ECOM48729.2019.8994884
• Yoon, K.-I., Jeong, T.-S., Kim, S.-C., & Lim, S.-C. (2023). Anonymizing at-home fitness: Enhancing privacy and motivation with virtual reality and try-on. Frontiers in Public Health, 11. https://doi.org/10.3389/fpubh.2023.1333776

Level: 

• Bachelor: Systematic Literature Review
• Master: Systematic Literature Review + Qualitative Research 

Contact: jana.lekscha(at)uni-bamberg.de 

Blockchain technology, known for its secure and decentralized structure, offers a potential solution for managing digital identities, ensuring transparency, security, and anonymity (Alabi, 2017; Raddatz et al., 2023). Despite its promise, adoption of blockchain identity systems remains in pilot stages, with many barriers to widespread use (Barbino, 2019). The research should explore the barriers and facilitators of adopting blockchain-based identity systems, focusing on how these systems are perceived by key stakeholders—governments, organizations, and users. By analyzing acceptance factors, it contributes to understanding the potential of blockchain. 

Literature: 

• Alabi, K. (2017). Digital blockchain networks appear to be following Metcalfe’s Law. Electronic Commerce Research and Applications, 24, 23–29.
• Barbino, V. H. (2019). Finding refuge: blockchain technology as the solution to the syrian refugee identification crises. 48 
• Liang, T.-P., Kohli, R., Huang, H.-C., & Li, Z.-L. (2021). What Drives the Adoption of the Blockchain Technology? A Fit-Viability Perspective. Journal of Management Information Systems, 38(2), 314–337. doi.org/10.1080/07421222.2021.1912915
• Raddatz, N., Coyne, J., Menard, P., & Crossler, R. E. (2023). Becoming a blockchain user: Understanding consumers’ benefits realisation to use blockchain-based applications. European Journal of Information Systems.

Level: 

• Bachelor: Systematic Literature Review
• Master: Systematic Literature Review + Qualitative Research 

Contact: jana.lekscha(at)uni-bamberg.de 

As the significance of social media continues to grow, phenomena such as cyberbullying and hate messages gain increasing prominence. According to a comparative study conducted by Bündnis gegen Cybermobbing e.V., approximately 11.5% of the German population experienced cyberbullying in 2021. This issue transcends the boundaries of the private sphere, extending its impact to the working environment. To comprehend the ensuing consequences, including depression, a thorough understanding of the dynamics within social media becomes paramount.
This thesis aims to investigate the dynamics of cyber abuse in social media using an analysis of existing relations. New datasets from platforms like Twitter will be meticulously collected, focusing on key terms, relevant time periods, and pivotal actors. Employing social media analytics methods, as outlined by Stieglitz et al. (2018), the study will comprehensively analyze and interpret social data, shedding light on the roles of actors, entities, and their relationships in the propagation of cyber abuse.

Literatur: 

• Topic discovery, data collection, and data preparation. International journal of information management, 39, 156-168.
• Beitzinger, F. & Leest U. (2021). Mobbing und Cybermobbing bei Erwachsenen: Eine empirische Bestandsaufnahme in Deutschland, Österreich und der deutschsprachigen Schweiz. [Online]. Available: www.buendnis-gegen-cybermobbing.de/mobbingstudie2021.html.
• D. K. Citron and H. Norton, “Intermediaries and hate speech: Fostering digital citizenship for our information age,” Bost. Univ. Law Rev., vol. 91, p. 1435, 2011.
• M.-A. Kaufhold, M. Bayer, and C. Reuter, “Rapid relevance classification of social media posts in disasters and emergencies: A system and evaluation featuring active, incremental and online learning, ” Inf. Process. Manag. , vol. 57, no. 1, pp. 1–32, 2020.
• K. Hartwig and C. Reuter, “TrustyTweet: An Indicator-based Browser-Plugin to Assist Users in Dealing with Fake News on Twitter,” 2019.

Level: Bachelor - Social Media Content Analysis

Kontakt: jana.lekscha(at)uni-bamberg.de 

The growing presence of cyberbullying and hate messages in digital environments has become alarmingly relevant. Protecting individuals from the harmful effects of these phenomena has become a pressing concern. Against this background, this research project aims to develop practical solutions to actively combat cyberbullying.

In this context, this research project aims not only to understand the dynamics of cyberbullying, but also to develop practical measures for containment and prevention. The research will focus on designing an innovative dashboard that integrates and visually represents AI-detected entities related to cyberbullying and hate messages. This dashboard will not only serve as a tool to respond to incoming reports, but also enable preventive action by identifying patterns and trends. Through a practice-oriented approach, this work intends to make a concrete contribution to the fight against cyberbullying and hate messages in digital spaces.

Literatur:

• Peffers, K., Tuunanen, T., Rothenberger, M. A., & Chatterjee, S. (2007). A Design Science Research Methodology for Information Systems Research. Journal of Management Information Systems, 24(3), 45–77.
• M.-A. Kaufhold, M. Bayer, D. Hartung, and C. Reuter , “Design and Evaluation of Deep Learning Models for Real- Time Credibility Assessment in Twitter,” in 30th International Conference on Artificial Neural Networks (ICANN2021), 2021, pp. 1–13, doi: doi.org/10.1007/978-3-030-86383- 8_32.
• K. Hartwig and C. Reuter, “TrustyTweet: An Indicator-based Browser-Plugin to Assist Users in Dealing with Fake News on Twitter,” 2019.
• M.-A. Kaufhold, Information Refinement Technologies for Crisis Informatics: User Expectations and Design Principles for Social Media and Mobile Apps. Wiesbaden: Springer Vieweg, 2021.

Level: Master - Design Science Research nach Peffers et al. (2007)

Kontakt: jana.lekscha(at)uni-bamberg.de 

Level: Master

Kontakt: jonas.rieskamp(at)uni-bamberg.de

In the realm of crisis communication, there is a growing concern, about the impact of social media-induced polarization (SMIP). Polarization is characterized by increasing divisions of opinions and attitudes and poses a significant threat. With the spread of information and misinformation on media, the risk of causing significant harm and widespread suffering among people increases. The goal of this thesis is to explore and understand the complexities of SMIP and examine what role social media platforms play in the increasing polarization. The problem at hand is exacerbated by information overload, where the constant exchange of information on these platforms reinforces existing beliefs, creating echo chambers and fostering a 'us versus them' mentality. The research focus is guided by Frame Theory, utilizing an algorithmic approach to delve into the causes of polarization and its impact on crisis communication.

Literatur:

• Qureshi, I., Bhatt, B., Gupta, S., & Tiwari, A. A. (2020). Causes, symptoms and consequences of social media induced polarization (SMIP). Information Systems Journal, 11.
• Cinelli, M., De Francisci Morales, G., Galeazzi, A., Quattrociocchi, W., & Starnini, M. (2021). The echo chamber effect on social media. Proceedings of the National Academy of Sciences, 118(9).
• Snow, D. A., Rochford Jr, E. B., Worden, S. K., & Benford, R. D. (1986). Frame alignment processes, micromobilization, and movement participation. American sociological review, 464-481.
• Johannessen, M. R. (2015). Please like and share! A frame analysis of opinion articles in online news. Pp. 15–26 in Lecture Notes in Computer Science.

Level:

Master

Kontakt: jonas.rieskamp(at)uni-bamberg.de

Contemporary AI applications are subject to biases, stemming from their training data. As a result, the outputs and decisions from AI applications are skewed towards the training data, lacking of fairness and inclusivity. AI ethics research submitted several design principles for AI application to enhance fairness. However, these principles are difficult to translate into practice, which leaves the fairness issues and resulting risks rather unaddressed. New approaches of AI risk management proposed the idea to “capture” Ai application in limited space in which it can act. This aims to contain harmful consequence (e.g., discrimination and unfairness) in a controllable environment. Yet, while the approach of ethics principles remain unfruitful, the capturing of AI contains negative consequence only retrospectively. A good solution, however, should act proactively. Considering the AI ethics issues as “IT failure” allows us to employ sociotechnical system perspective, which investigated solution for issues emerging from IT projects. This thesis will summarise current AI ethics issues and categorises them according to sociotechnical system perspective. Upon successful categorisation, proactive solutions will be derived and synthesised. 

Literatur:

• Bostrom, R. P., & Heinen, J. S. (1977). MIS Problems and Failures: A Socio-Technical Perspective. Part I: The Causes. MIS Quarterly, 1(3), 17–32. https://doi.org/10.2307/248710
• Bostrom, R. P., & Heinen, J. S. (1977). MIS Problems and Failures: A Socio-Technical Perspective, Part II: The Application of Socio-Technical Theory. MIS Quarterly, 1(4), 11–28. https://doi.org/10.2307/249019
• Asatiani, A., Malo, P., Nagbøl, P. R., Penttinen, E., Rinta-Kahila, T., & Salovaara, A. (2021). Sociotechnical Envelopment of Artificial Intelligence: An Approach to Organizational Deployment of Inscrutable Artificial Intelligence Systems. Journal of the Association for Information Systems, 22(2), 325–352. https://doi.org/10.17705/1jais.00664
• Mirbabaie, M., Brendel, A. B., & Hofeditz, L. (2022). Ethics and AI in Information Systems Research. Communications of the Association for Information Systems, 50(1), 726–753. https://doi.org/10.17705/1CAIS.05034
• Laine, J., Minkkinen, M., & Mäntymäki, M. (2025). Understanding the Ethics of Generative AI: Established and New Ethical Principles. Communications of the Association for Information Systems, 56(1). https://aisel.aisnet.org/cais/vol56/iss1/7
• Mittelstadt, B. (2019). Principles alone cannot guarantee ethical AI. Nature Machine Intelligence, 1(11), 501–507. https://doi.org/10.1038/s42256-019-0114-4 

Level:

• Bachelor: Systematic Literature Review

Kontakt: jonas.rieskamp(at)uni-bamberg.de

The rise of agentic IS systems—intelligent information systems that can autonomously perform tasks, learn from interactions, and delegate actions—has fundamentally reshaped human-technology collaboration (Baird & Maruping, 2021). Large Language Models (LLMs), such as GPT-based architectures, are increasingly being integrated into multi-agent systems, where they serve as decision-making entities, negotiators, and collaborative agents in distributed environments (Göldi & Rietsche, 2024). These LLM-based agents demonstrate emergent coordination capabilities, enabling complex problem-solving and adaptive responses across dynamic scenarios (Guo et al., 2024). However, the effectiveness of such systems also depends on how autonomy and control are balanced in human-AI interactions. Recent research highlights that user acceptance and system efficiency are influenced by whether task delegation is initiated by users or the AI system itself, underscoring the importance of designing adaptive delegation mechanisms (Adam et al., 2024).

Despite their potential, LLM-based multi-agent systems face challenges such as ensuring alignment with human goals, managing inter-agent communication, and mitigating unintended biases or behaviors. This Bachelor’s thesis aims to systematically review the existing body of research on LLM-based multi-agent systems, synthesizing their theoretical foundations, applications, and ongoing challenges. Furthermore, the thesis will explore emerging opportunities for information systems research by examining how LLM-based multi-agent systems contribute to new forms of organizing, decision-making, and collaboration in digital environments.

Literatur:

• Baird, A., & Maruping, L. M. (2021). The Next Generation of Research on IS Use: A Theoretical Framework of Delegation to and from Agentic IS Artifacts. MIS Quarterly, 45(1), 315–341. https://doi.org/10.25300/MISQ/2021/15882
• Göldi, A., & Rietsche, R. (2024). Making Sense of Large Language Model-Based AI Agents. Proceedings of the Forty-Fifth International Conference on Information Systems. https://aisel.aisnet.org/icis2024/aiinbus/aiinbus/16
• Guo, T., Chen, X., Wang, Y., Chang, R., Pei, S., Chawla, N. V., Wiest, O., & Zhang, X. (2024). Large language model based multi-agents: A survey of progress and challenges. Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, 8048–8057. doi.org/10.24963/ijcai.2024/890
  Adam, M., Diebel, C., Goutier, M., & Benlian, A. (2024). Navigating autonomy and control in human-AI delegation: User responses to technology- versus user-invoked task allocation. Decision Support Systems, 180, 114193. https://doi.org/10.1016/j.dss.2024.114193
• Webster, J., & Watson, R. T. (2002). Analyzing the Past to Prepare for the Future: Writing a Literature Review. MIS Quarterly, 26(2), xiii–xxiii.
   

Level:

• Bachelor

Kontakt: jonas.rieskamp(at)uni-bamberg.de

Remote work practices have become increasingly prevalent in organisations. Yet, it remains puzzling why remote work at scale, that is, remote organising, creates substantive challenges for transforming organisations, while perennial all-remote organisations seem to thrive with it. Many all-remote organisations openly share and promote their work processes through remote work ‘handbooks,’ ‘guidelines,’ and ‘manifestos.’ The goal of this thesis is to qualitatively analyse the ‘handbooks,’ ‘guidelines,’ and ‘manifestos’ to improve our understanding of remote organising. 

Literatur:

• Brünker, F., Marx, J., Mirbabaie, M., & Stieglitz, S. (2023). Proactive digital workplace transformation: Unpacking identity change mechanisms in remote-first organisations. Journal of Information Technology, 0(0), 1-19. https://doi.org/10.1177/02683962231219516 

• Choudhury, P. (Raj)., Foroughi, C., & Larson, B. (2021). Work-from-anywhere: The productivity effects of geographic flexibility. Strategic Management Journal, 42(4), 655–683. https://doi.org/10.1002/smj.3251

• Rhymer, J. (2022). Location-Independent Organizations: Designing Collaboration Across Space and Tme. Administrative Science Quarterly, 68(1), 1-43. https://doi.org/10.1177/00018392221129175

Level: Master

Kontakt: j.marx(at)unimelb.edu.au

The relationship between humans and AI systems is of central importance as AI becomes more and more integrated into our everyday lives. This relationship not only influences the way we work and communicate, but also our decision-making and our trust in technologies. A deep understanding of these dynamics can help to develop AI systems that are ethical, transparent and user-friendly. This topic is particularly relevant as it examines the interface between technology and human behavior and thus provides important insights for the design of information systems.
A thesis could focus on investigating trust building between users and AI systems. This could be done through surveys and experiments testing different interaction designs to find out which factors strengthen or weaken user trust.
You are also welcome to contact me with your own thesis ideas on this topic. 

Literatur:

• Pal, D., Vanijja, V., Thapliyal, H. & Zhang, X. (2023). What affects the usage of artificial conversational agents? An agent personality and love theory perspective. Computers in Human Behavior, 145, 107788.
  doi.org/10.1016/j.chb.2023.107788
• Song, X., Xu, B. & Zhao, Z. (2022b). Can people experience romantic love for artificial intelligence? An empirical study of intelligent assistants. Information & Management, 59(2), 103595.
  doi.org/10.1016/j.im.2022.103595

Level: Bachelor

Kontakt: milad.mirbabaie(at)uni-bamberg.de

AI ethics is essential because it ensures that AI is in line with human values and rights, promotes trust through transparency and accountability, protects against abuse and harm through guidelines for fair and safe use, and supports justice by minimising discrimination and prejudice. It also emphasises privacy through responsible data handling and supports security by identifying and mitigating risks and threats.
A thesis topic could be an investigation how the use of AI technologies affects the digital divide between different social groups. To do this, an analysis of the accessibility and usability of AI systems for different population groups can be carried out. 
You are also welcome to contact me with your own thesis ideas on this topic. 

Literatur: 

• Mirbabaie, M., Brendel, A. B. & Hofeditz, L. (2022). Ethics and AI in Information Systems Research. Communications Of The Association For Information Systems, 50(1), 726–753. 
  https://doi.org/10.17705/1cais.05034
• Floridi, L. & Cowls, J. (2021). A unified framework of five principles for AI in society. In Philosophical studies series (S. 5–17). 
  doi.org/10.1007/978-3-030-81907-1_2

Level: Bachelor

Kontakt: milad.mirbabaie(at)uni-bamberg.de

Black Friday presents a major opportunity for businesses but with thousands of companies competing for consumer attention, standing out is a challenge. Companies must develop effective strategies to capture interest and drive sales, particularly on social media, where engagement can make or break a campaign. The question is: How can businesses maximize their impact on platforms like Twitter surrounded by the noise of countless other promotions? The goal is to analyze and understand different Black Friday company strategies within the Social Commerce context. How do companies utilize social media platforms, such as Twitter, to attract customers? These strategies may vary in terms of content and promotional techniques. Key areas of focus include:

• How do companies structure their Black Friday campaigns on Twitter?

• What are the key elements of an effective social media strategy for Black Friday?

• How do companies measure the success of their social media strategies on Black Friday?

• What challenges arise in managing social media campaigns?

Methodology: This study will involve interviews with social media managers from various companies, content analysis, and other relevant research methods.

Literature: 

• Lin, X., & Wang, X. (2023). Towards a model of social commerce: Improving the effectiveness of e-commerce through leveraging social media tools based on consumers’ dual roles. European Journal of Information Systems, 32(5), 782–799. https://doi.org/10.1080/0960085X.2022.2057363
• Leong, L.-Y., Hew, T. S., Ooi, K.-B., Hajli, N., & Tan, G. W.-H. (2024). Revisiting the social commerce paradigm: The social commerce (SC) framework and a research agenda. Internet Research, 34(4), 1346–1393. https://doi.org/10.1108/INTR-08-2022-0657

• Li, F., Larimo, J., & Leonidou, L. C. (2021). Social media marketing strategy: Definition, conceptualization, taxonomy, validation, and future agenda. Journal of the Academy of Marketing Science, 49(1), 51–70. https://doi.org/10.1007/s11747-020-00733-3

Level: Bachelor oder Master

Kontakt: truong-ngoc(at)information-systems.org

In information systems (IS) research, extracting meaningful insights from short-text data, such as Twitter tweets, remains a challenge. Traditional topic modeling approaches often struggle with the brevity and informal nature of such texts, potentially limiting their usefulness for IS theory development. Given the rise of newer techniques like BERTopic, Top2Vec, and large language models (LLMs), it is crucial to understand how these methods compare in generating relevant topics.

In the IS literature, various topic modeling methods have been used to investigate social and organizational phenomena. It would be interesting to compare these methods, particularly in their application to short-text data such as Twitter tweets. Specifically, how do the topics generated by BERTopic, Top2Vec, (author-pooled) LDA, and (zero-shot) LLMs/GenAI differ in terms of their relevance for IS theory construction? Alternative methods can also be considered.

Methodology: Evaluation of Twitter data using AI/ML methods, validated through comparison with human coding.

Literature: 

• Berente, N., Seidel, S., & Safadi, H. (2019). Research Commentary—Data-Driven Computationally Intensive Theory Development. Information Systems Research, 30(1), 50–64. https://doi.org/10.1287/isre.2018.0774
• Kishore, S., Sundaram, D., & Myers, M. D. (2024). A temporal dynamics framework and methodology for computationally intensive social media research. Journal of Information Technology, in press. https://doi.org/10.1177/02683962241283051
• Miranda, S., Berente, N., Seidel, S., Safadi, H., & Burton-Jones, A. (2022). Computationally Intensive Theory Construction: A Primer for Authors and Reviewers. MIS Quarterly, 46(2), iii–xviii.
• Yang, Y., & Subramanyam, R. (2023). Extracting Actionable Insights from Text Data: A Stable Topic Model Approach. MIS Quarterly, 47(3), 923–954. https://doi.org/10.25300/MISQ/2022/16957

Level: Bachelor oder Master

Kontakt: truong-ngoc(at)information-systems.org

Many state-of-the-art AI tools, such as BERTopic for topic modeling, are often perceived as black boxes because it is unclear how these models arrive at their results. This lack of transparency can discourage researchers from using AI to analyze data, such as Twitter posts. Explainable AI (XAI) techniques could help address this issue. 

This study will explore whether and how XAI techniques (e.g., SHAP, LIME) can be integrated into transformer-based models for topic modeling or sentiment analysis. The goal is to assess the impact of these methods on the trust and efficiency of IS researchers.

Methodology: Application of AI for coding Twitter data with XAI, followed by human validation.

Literature: 

• Stoffels, D., Faltermaier, S., Strunk, K. S., & Fiedler, M. (2024). Guiding computationally intensive theory development with explainable artificial intelligence: The case of shapley additive explanations. Journal of Information Technology, https://doi.org/10.1177/02683962241289597
• Fernández-Loría, Carlos; Provost, Foster; and Han, Xintian. 2022. "Explaining Data-Driven Decisions made by AI Systems: The Counterfactual Approach," MIS Quarterly, (46: 3) pp.1635-1660. https://doi.org/10.25300/misq/2022/16749

Level: Bachelor oder Master

Kontakt: truong-ngoc(at)information-systems.org