Evaluation of literature by professional and layperson critics: A digital and literary sociological analysis of evaluative talk of literature through the prism of literary prizes (2007-2017)
Aspect-based Sentiment Analysis (ABSA) of German-language social media data related to the Ingeborg-Bachmann-Preis (Tage der deutschsprachigen Literatur TDDL), a German-Language Literature festival. For more background information, please consult the project website
We collected tweets published about the TDDL between 2007 and 2019 using the freely available scraping tool “OMGOT3” and a predefined list of topical hashtags.
All social media data were/are publicly accessible and have been anonymized during pre-processing. The collected data are used for non-commercial, academic purposes only. No information on individual users, apart from their handle and the number of likes and retweets, was collected – in keeping with the social media platform’s API access terms. Academics who wish to replicate our research can collect the data using the provided tweet IDs, which give access to the original tweet content if it is still publicly available. More information on Twitter’s policy and recently expanded access for academic research is to be found here.
Please note that tweets are constantly being created or removed and that some profiles may no longer be public. As a consequence, your collected corpus may differ slightly from ours.
Requirements:
- Python 3.6 +
- pip
- Jupyter Notebook
Setup:
-
git init -
git clone https://github.com/gnmarten/talklitmining-CLIN31 -
cd talklitmining-CLIN31 -
pip install -r requirements.txt -
jupyter notebook
The repository is structured as followed:
- data/ : This directory contains all the annotated CSVs, and stores all the intermediary files while preprocessing
- preprocessing.ipynb : Contains all the functions and code to download Tweet texts, and setup data for training for the model
- train_and_eval.ipynb : Trains model from scratch and runs evaluation on the test sets
To replicate the experiments, run preprocessing.ipynb and train_and_eval.ipynb sequentially. Note: Add Twitter API developer keys, at the top of the file preprocessing.ipynb to get the code running!
You can read the paper here. Please cite us if you found this useful.
De Greve, L., Singh, P., Van Hee, C., Lefever, E., & Martens, G. (2022). Aspect-based Sentiment Analysis for German: Analyzing “Talk of Literature” Surrounding Literary Prizes on Social Media. Computational Linguistics in the Netherlands Journal, 11, 85–104. Retrieved from https://www.clinjournal.org/clinj/article/view/142
@article{greve_aspect-based_2021,
title = {Aspect-based {Sentiment} {Analysis} for {German}: {Analyzing} “{Talk} of {Literature}” {Surrounding} {Literary} {Prizes} on {Social} {Media}},
volume = {11},
copyright = {Copyright (c) 2022},
issn = {2211-4009},
shorttitle = {Aspect-based {Sentiment} {Analysis} for {German}},
url = {https://www.clinjournal.org/clinj/article/view/142},
language = {en},
urldate = {2022-03-21},
journal = {Computational Linguistics in the Netherlands Journal},
author = {Greve, Lore De and Singh, Pranaydeep and Hee, Cynthia Van and Lefever, Els and Martens, Gunther},
year = {2021},
pages = {85--104},
url = {https://www.clinjournal.org/clinj/article/view/142}
}
Open an issue for bugs or trouble with running the scripts.
For further questions related to the code contact: