Text-as-Data

Framing Democracy: Identifying Autocratic Anti-Democratic Propaganda Using Word Embeddings

Patrick Chester (New York University)

Abstract: There is substantial empirical evidence that indicates that democracy can spread between countries through observational learning. But do autocracies try to bias learning against...

Read more about Framing Democracy: Identifying Autocratic Anti-Democratic Propaganda Using Word Embeddings

Decoding Propaganda Slogans in China: Reading Between the Lines Using Word Embeddings

Yin Yuan (University of California, San Diego)

Abstract: Propaganda slogans in China (a.k.a. “catchphrases” or “tifa”) are widely believed to be artifacts of propaganda aimed at indoctrinating the general public that convey little substantive political or policy information. This paper intends to show instead that these...

Read more about Decoding Propaganda Slogans in China: Reading Between the Lines Using Word Embeddings

Religiosity and Secularism: A Text-as-Data Approach to Recover Jihadist Groups' Rhetorical Strategies

Luwei Ying (Washington University in St. Louis)

*Award for Best Graduate Student Poster - Applications*

Abstract: Radical Islamists as the major force of the current "wave" of terrorism pursue impact, not only attacks. Scholars, however, for decades have almost exclusively focused...

Read more about Religiosity and Secularism: A Text-as-Data Approach to Recover Jihadist Groups' Rhetorical Strategies

Joint Image-Text Classification Using a Transformer-Based Architecture

Patrick Wu and Walter R. Mebane Jr. (University of Michigan)

Abstract: The use of social media data in political science is now commonplace. Social media posts such as Tweets are usually multimodal, comprising...

Read more about Joint Image-Text Classification Using a Transformer-Based Architecture

Rigorous Subjectivity: Demystifying and Improving Human Coding With Statistical Models

Matthew Tyler (Stanford University)

Abstract: Researchers are often tasked with applying subjective or contested labels to objects such as text and images. For example, researchers might hire coders to label the ideological slant of news articles. I show how two typical coding workflows in political science, traditional small-team...

Read more about Rigorous Subjectivity: Demystifying and Improving Human Coding With Statistical Models

Finding Social Media Trolls: Dynamic Keyword Selection Methods for Rapidly Evolving Online Debates

Maya Srikanth, Anqi Liu, Nicholas Adams-Cohen, Anima Anandkumar and R. Michael Alvarez (California Institute of Technology)

Abstract: Online harassment is a significant social problem. Prevention of online harassment requires rapid detection of harassing, offensive, and negative social media posts. In this paper, we propose the use of word embedding...

Read more about Finding Social Media Trolls: Dynamic Keyword Selection Methods for Rapidly Evolving Online Debates

The Politics of Science: Evidence From 19th-Century Public Health

Casey Petroff (Harvard University)

Abstract: How do governments decide between protecting public health and protecting the economy when a new disease threat emerges? I study this question using evidence from cholera epidemics in the 19th century. In the face of this new threat to public health, professional opinion was divided between...

Read more about The Politics of Science: Evidence From 19th-Century Public Health

Who Do You Think You’re Fooling? Examining the Internal Russian Disinformation Campaign

Sean Norton (University of North Carolina, Chapel Hill)

Abstract: While an extensive literature in both the academic and popular presses has examined the Russian state’s use of Twitter to influence the 2016 US...

Read more about Who Do You Think You’re Fooling? Examining the Internal Russian Disinformation Campaign

The Moral Narrative of the European Sovereign Bond Crisis

Nicola Nones (University of Virginia)

Abstract: In this paper, I take a first step towards assessing if and to what extent the debt crisis has given rise to a moral narrative that starkly divides virtuous Northern European countries on the one side, and spendthrift, lazy Southern European ones on the other side. Such moral...

Read more about The Moral Narrative of the European Sovereign Bond Crisis

Beyond Topics: Semi-Supervised Learning for Texts From a Measurement Perspective

Shiyao Liu (Massachusetts Institute of Technology)

Abstract: This project proposes a new methodological framework to use text data as a measurement in political science. Despite the abundance of text data available nowadays, conversion of text data into a measurement for a political concept remains a challenge that prevents...

Read more about Beyond Topics: Semi-Supervised Learning for Texts From a Measurement Perspective

Diffusion of Cybersecurity Policies

Nadiya Kostyuk (University of Michigan, Ann Arbor)

Abstract: One of the most important developments of the last two decades has been the spread of national cybersecurity policies that affect millions of people globally. Yet, researchers know relatively little about this phenomenon. I study cybersecurity policy diffusion...

Read more about Diffusion of Cybersecurity Policies

The Spread of Promotion of Political Violence on Twitter

Taegyoon Kim (Pennsylvania State University)

Abstract: Although social media contribute to political participation by enabling citizens to freely express and exchange political opinions, increasing concerns are raised about the...

Read more about The Spread of Promotion of Political Violence on Twitter

Paragraph-Citation Topic Models for Corpora With Citation Networks

ByungKoo Kim, Yuki Shiraito and Saki Kuzushima (University of Michigan)

Abstract: Social scientists often analyze a corpus with a citation network among its documents, such as the corpus of the U.S. Supreme Court decisions. Existing topic models for document networks assume that the topic of a citation is...

Read more about Paragraph-Citation Topic Models for Corpora With Citation Networks

Violation of Bright Lines, Term Limit Evasion and Information Control

JunHyeok Jang (University of California, Merced)

Abstract: Constitutional “bright lines” are generally thought to serve as an important guard against democratic breakdown, because violations of these “bright line” institutions provide a focal point that facilitates mass coordination against a leader. In countries with term...

Read more about Violation of Bright Lines, Term Limit Evasion and Information Control

Extracting Political Events From Text Using Grammatical Dependency Parsing and Machine Learning

Andrew Halterman (Massachusetts Institute of Technology)

Abstract: This paper introduces a method that automatically extracts political events from text using grammatical parsing and machine learning. Much of the scientifically useful information about what political actors are doing is locked away in text. To extract this...

Read more about Extracting Political Events From Text Using Grammatical Dependency Parsing and Machine Learning

Pages