This repository contains the code to replicate the results in "Towards a large scale analysis of claims: developing a machine learning method for detecting and classifying politicians’ claims of representation".

To run this code, please install the necessary packages using the requirements.txt file. The notebooks that are used to create the final annotations can be found in "notebooks_to_run". For some steps, there are .ipynb and .py files available, either one can be used. The initial number in the file name represents when this file should be ran in the sequence (for instance, the output of the 0_preprocessing.ipynb step will be used in 1_object.ipynb). Because the code and files that are imported in the notebooks are too large, we don't upload them in this repository. Instead, they can be requested by sending an email to [email protected] . Paste this folder as "code" in the repository if you want to run the notebooks.

This code is made to detect and classify different claims of representation in Dutch social media data. The data is not currently available, but this code can be used on new data. The necessary information to run this code as-is are: ID-column, text, name of the person that posted it, date of post. The folder "data" is currently empty. Here, you can put your excel file, and this will also be used to save predictions of the notebooks.

If you use (parts of) this code, please cite our paper: @article{gevers2024towards, title={Towards a large scale analysis of claims: developing a machine learning method for detecting and classifying politicians’ claims of representation}, author={Gevers, Ine and De Mulder, August and Daelemans, Walter}, journal={Journal of Computational Social Science}, pages={1--45}, year={2024}, publisher={Springer} }