TNA Fellow Julia Neugarten
Catching Feelings: Aspect-Based Sentiment Analysis for Fanfiction about Greek Myth
Fanfiction – stories inspired by existing stories, written by and for fans and published online for free – often revolves around emotional trajectories. Emotional highs and lows are frequently central to fanfiction plots. This focus on emotion makes fanfiction interesting to analyze with computational sentiment analysis methods.
The aim of ‘Catching Feelings’ is to apply aspect-based sentiment analysis (ABSA) to a corpus of over 5.000 fanfiction-texts about Greek myth, as well as a corpus of reviews for those stories. This analysis fits within the PhD-project ‘Anchoring and Innovating Classical Motifs in Fanfiction’, which analyzes how source material from Greco-Roman Antiquity is transformed in contemporary online fanfiction.
Using ABSA, the aim is to answer the following research questions:
1. How does fanfiction present emotional engagement with Greek myth?
2. Which aspects of this source material are loved, enjoyed, and appreciated?
By approaching sentiment both as a property of texts and as a phenomenon that arises in the encounter between texts and readers, the project addresses an important theoretical matter for the application of sentiment analysis to narrative: where does the emotional impact of a story reside? Can it be measured in the text itself, or is it located in reader responses?
In this video, CLS INFRA TNA Fellow Julia Neugarten of Radboud University, Nijmegen, speaks about the project: ‘Catching Feelings: Aspect-Based Sentiment Analysis for Fanfiction about Greek Myth’
CLS INFRA TNA Fellow Julia Neugarten gives a presentation on her project “Catching Feelings: Aspect-Based Sentiment Analysis for Fanfiction about Greek Myth”. Julia and a team of researchers at GhentCDH & Lt3 are mapping the ways fanfiction-readers evaluate the stories they read in the Greek mythology fandom on popular fanfiction-website “Archive of Our Own”. Using aspect-based sentiment analysis (ABSA), Julia and her colleagues examine how several aspects of fanfiction about Greek myth are evaluated by commenters. In this presentation, Julia will report on the processes of data collection and annotation for the project, introduce the different aspects, describe the pipeline for aspect-based sentiment analysis and present some preliminary results. Julia Neugarten is currently at General Cultural Sciences at Radboud University Nijmegen within the Anchoring Innovation project.