Comparison of faction detection methods in application to Ukrainian parliamentary data

Abstract

In this study we compare two general methods of faction detection from Ukrainian Parliamentary roll call data, MacRae’s method and Gower’s method. Both methods were adapted to the special voting procedures and patterns of the Ukrainian Parliament, such as its nonbinary voting scheme. Our analysis shows that each method is viable for faction detection individually, and that both can be used in tandem for results with higher confidence. Viability was demonstrated through the construction of the cooperation network between official parties, and by listing key parliamentarians based on their centrality in the factionized voting network. While the party-party network is intuitive, it was found that a pair of key actors in one faction were from opposing parties.

Publication
Late-Breaking Paper In SBP-BRiMS 2018
Tom Magelinski, PhD
Tom Magelinski, PhD
Senior Data Scientist - Information Extraction and Generative AI

I build AI systems that help domain experts understand vast amounts of data through state-of-the-art techniques from natural language processing, generative modeling, graph ML, and network science. I’m particularly interested researching and developing techniques to combine NLP and graph-based approaches to capture complex relationships in unstructured data.

Related