Computational Text Analysis

Institutional Grammar Research Initiative (IGRI) affiliates are engaged in research focusing on the development and application of computational text analysis and supervised machine learning approaches for evaluating institutions based on the Institutional Grammar.

Below is a list of projects relating to this research theme currently being pursued by IGRI affiliates.


Project Title: Machine Coding of Policy Texts with the Institutional Grammar

Project Description: Applying the Grammar of Institutions in a particular research setting is resource- and time-intensive, precipitating concerns over whether it might ever enjoy widespread use. To address this, our team of scholars is developing automated approaches and open-source tools for coding institutions based on the Grammar. Preliminary work, applying insights from computational linguistics and natural language processing, was recently published in Public Administration as part of a special issue on methodological advances related to the Grammar. Currently, the team is continuing the development of the approach, while also preparing an R package to increase the accessibility of the Grammar for other scholars.

Project Team: Douglas Rice, Saba Siddiki, Seth Frey, Hyoung Keun (Jay) Kwon, Adam Sawyer

Funding Sources: UMass Amherst Institute for Social Science Research (ISSR) Sustainability Seed Grant, Syracuse University Collaboration for Unprecedented and Excellence (CUSE) Grant, National Science Foundation Research Coordination Network Grant.