Context

Menstrual and reproductive health affects nearly half the global population, yet awareness and literacy remain critically low. Despite 1.8 billion people menstruating worldwide, stigma and symptom normalization lead to widespread neglect of conditions like heavy bleeding, irregular cycles, or chronic pelvic pain. A striking example is endometriosis, affecting 1 in 10 menstruating individuals and taking up to 10 years on average to diagnose – often worsening symptoms and impairing fertility. Closing the menstrual health literacy gap is essential for improving quality of life and timely care.

While digital health technologies, especially menstrual tracking apps, have shown promise in symptom management (e.g., for PMS or PMDD), their educational potential remains underexplored. In particular, it is still unclear how personalized menstrual health content can foster health literacy and awareness among diverse user groups (e.g., teenagers, individuals with menstrual issues, or individuals in supportive roles).

Scope of the thesis

This Master's thesis aims to investigate the potential of Large Language Models (LLMs) in improving menstrual health literacy. The research will explore how LLMs can benefit both individual users (menstruating individuals across various life stages, including pre-menopause and menopause) and the broader societal context (non-menstruating individuals, such as men). The overarching goal is to develop and evaluate a system that provides personalized menstrual health content, primarily leveraging open-source data from social media. This thesis will focus on the following core activities:

• Literature Review & Research Question
• Data Collection & Preprocessing: Scrape and analyze social media (e.g., TikTok/Instagram com-ments or Reddit threads) data (e.g., questions, dis-cussions, misconceptions) related to menstrual health content
• Designing LLM-based Content Recommen-dation: Use a Large Language Model (e.g., GPT) to generate personalized educational responses. Im-plement mechanisms to tailor these responses to user context or segment (e.g., teenagers, people with menstrual disorders, partners)
• Evaluation: Assess the alignment and relevance of LLM-generated recommendations

Requirements

We are seeking a highly motivated Master's student with a strong interest in digital health and natural language processing with programming skills in Python, familiarity with or willingness to learn LLM APIs, ability to clearly articulate technical concepts and research findings in written and oral form.

What We offer

Supervision and mentoring by an interdisciplinary team at the intersection of digital health and gender equity in collaboration with KIT, University of St. Gallen and ETH Zürich.

Related Work

Sou, D., Fuchs, M., Meegahapola, L., Peintner, A., Jutzeler, C.R., Kowatsch, T., Ivanova, O., Nißen, M., How self-tracking and engagement with personalized health content shape self-reported menstrual health experiences in a women’s health app, Research Square, 8 July 2025, Preprint, 10.21203/rs.3.rs-7019879/v1.

Tumescheit, C., Sou, D., Nißen, M.K., Kowatsch, T., Hastings, J., Large Language Models Reveal Menstruation Experiences and Needs on Social Media, 20th World Congress on Medical and Health Informatics (Medinfo), Taipei, 9-13 August 2025 (forthcoming).