Maddalena Ghiotti
A Data-Driven Investigation of Potentially Harmful Diet-Related Content on YouTube: Detection and User Engagement Analysis.
Rel. Tatiana Tommasi, Daniela Paolotti, Yelena Aleksandrovna Mejova. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Matematica, 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (7MB) | Preview |
Abstract
In recent years, social media have become popular sources of information, including in the context of nutrition and weight loss. However, the spread of low-quality content that emphasizes physical appearance poses a risk to viewers’ self-esteem, body image, and body satisfaction, and in the most severe cases, it may contribute to the development of disordered eating behaviors (DEBs) or, potentially, eating disorders (EDs). This study analyzes 3129 YouTube videos about diet and weight loss, considering titles, descriptions, transcripts, and metadata. The analysis aims to explore the relationship between these metadata, video topics, and the level of risk associated with each video, to facilitate the detection of harmful material and inform moderation practices.
We quantify the risk using 23 principles derived from three questionnaires
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
