Sentiment Analysis of Fanfiction Texts

This data repository contains derived data for the paper Happily Ever After: Comparing Sentiment Arcs in Emotionally-Inflected Fanfiction Genres Across Fandoms by Julia Neugarten, Pascale Feldkamp, Mia Jacobsen and Yuri Bizzoni. It has been accepted as a long paper to Computational Humanities Research (CHR) 2025. See the readme-file for the paper's abstract.

This data repository contains the file anonymized_merged_sentiment_data.csv, which contains derived data for all 12.199 stories in the dataset of fanfiction analyzed in Happily Ever After. Because of copyright, the fanfiction itself cannot be shared. All fanfiction was collected from popular fanfiction-platform Archive of Our Own (AO3) in accordance with their terms of service.

The dataset we used for this paper is a combination of two datasets previously described in publications by Mia Jacobsen and Julia Neugarten respectively, and cited in the readme-file of this repository. The file anonymized_merged_sentiment_data.csv also includes relevant metadata per story and sentiment scores and statistics calculated with our two methods: the Syuzhet-package and the BERT model.

The codebook included in this repository contains more detailed description of the data provided in the csv-file. In addition, all software used to conduct the sentiment analysis for our CHR paper is available on Github.