Aspect Based Sentiment Analysis and User Generated Book Reviews for Arabic Literature on Goodreads (ABSA-UGBR)

Aspect Based Sentiment Analysis and User Generated Book Reviews for Arabic Literature on Goodreads

Acronym: ABSA-UGBR
Partners: Christian Junge (Arabic Studies), Alena Strohmaier (Media Studies), Johanna Fuchs (Near and Middle Eastern Studies)
Funding: Creative Space - Marburg University
Status: Ongoing
Start: 2025
End: 2025

Social media platforms such as YouTube, TikTok, and Goodreads offer readers from all over the world the opportunity to express their feelings, opinions, and reviews of books. These so-called user-generated book reviews (UGBR) provide a vast amount of reading reports and reader interpretations, offering historically unique insights into literary reception and recent reading practices in the digital age. While literary studies have recently recognized the relevance of this data and are conducting qualitative analyses of small corpora, a systematic computer-based quantitative and qualitative analysis of larger corpora, for example using Aspect Based Sentiment Analysis (ABSA), a sentiment analysis that brings together different aspects of statements and evaluations, is still largely lacking.

This project aims to test the technical feasibility and interdisciplinary relevance of Aspect Based Sentiment Analysis for the evaluation of user-generated book reviews in English and Arabic on the social platform Goodreads.