Deep Learning for Political Science

Abstract

IntroductionPolitical science, and social science in general, have traditionally been using computational methods to study areas such as voting behavior, policy making, international conflict, and international development. More recently, increasingly available quantities of data are being combined with improved algorithms and affordable computational resources to predict, learn, and discover new insights from data that is large in volume and variety. New developments in the areas of machine learning, deep learning, natural language processing (NLP), and, more generally, artificial intelligence (AI) are opening up new opportunities for testing theories and evaluating the impact of interventions and programs in a more dynamic and effective way.Applications using large volumes of structured and unstructured data are becoming common in government and industry, and increasingly also in social science research.

Text mining for social science – The state and the future of computational text analysis in sociology

“…A. Macanovic algorithmic bias stemming from the fact that these algorithms are designed by humans and learn patterns from human-generated data (Chatsiou and Mikhaylov 2020;Waseem 2016). For example, research has shown that hate speech detection systems that use supervised classification learn and reproduce biases from manually coded data they received as input, effectively discriminating against the groups they were initially designed to protect (see, for example, Davidson and Bhattacharya 2020).…”

“…In the last several years, new developments in deep learning have further improved the quality of language representations. Resembling human information processing, deep learning algorithms iteratively transform the initial (textual) information input into more and more abstract representations (Chatsiou and Mikhaylov 2020;Minaee et al, 2021). Recently developed transformer models are capable of assessing the plurality of contexts in which the same word can occur in text (Vaswani et al, 2018), thus capturing the nuances of meaning even better than simpler word embedding and deep learning models.…”

POLITICS: Pretraining with Same-story Article Comparison for Ideology Prediction and Stance Detection

“…Given the importance of ideology prediction and stance detection tasks in political science (Thomas et al, 2006;Wilkerson and Casas, 2017;Chatsiou and Mikhaylov, 2020), we conduct extensive experiments on a wide spectrum of datasets with 11 tasks ( §5.1). We then compare with both classical models and prior PLMs ( §5.2), and among our model variants ( §5.3).…”

Sosyal Bilimlerde Büyük Veri Analitiği, Yapay Zeka ve Makine Öğreniminin Kullanımı

“…Derin öğrenme, sosyal bilimler araştırmalarında oy verme davranışı, politika oluşturma, çatışma ve kalkınma gibi alanlarda kullanılmaktadır. Örneğin Chatsiou ve Mikhaylov (2020) tarafından yürütülen çalışmada derin öğrenmenin siyaset bilimi metinlerinden bilgi çıkarma ve örüntü tanımlama potansiyeli gözden geçirilmektedir (Chatsiou ve Mikhaylov, 2020). Derin öğrenme, psikoloji alanı için de yeni araştırma yöntemleri sağlamakta ve yeni fikirler sunmaktadır.…”