Political Advertising Dataset: the use case of the Polish 2020 Presidential Elections

TL;DR

This paper introduces the first publicly available dataset of Polish political tweets, annotated for campaign categories, and demonstrates its use in training a neural tagger to analyze political advertising during the 2020 presidential elections.

Contribution

It provides a novel annotated dataset for Polish political ads and applies neural models for categorization, enabling new research in political social media analysis.

Findings

Achieved 0.65 inter-annotator agreement (Cohen's kappa)

Neural tagger reached 70% F1 score on the dataset

Initial analysis of Polish 2020 Presidential Elections on Twitter

Abstract

Political campaigns are full of political ads posted by candidates on social media. Political advertisements constitute a basic form of campaigning, subjected to various social requirements. We present the first publicly open dataset for detecting specific text chunks and categories of political advertising in the Polish language. It contains 1,705 human-annotated tweets tagged with nine categories, which constitute campaigning under Polish electoral law. We achieved a 0.65 inter-annotator agreement (Cohen's kappa score). An additional annotator resolved the mismatches between the first two annotators improving the consistency and complexity of the annotation process. We used the newly created dataset to train a well established neural tagger (achieving a 70% percent points F1 score). We also present a possible direction of use cases for such datasets and models with an initial analysis of the Polish 2020 Presidential Elections on Twitter.