CCP’s Concept Integration in Comparative Law project has launched two new digital tools to aid researchers working with large corpora like constitutions and court rulings.
The team’s suite of open-source concept analysis tools includes a new constitution comparison tool (to assess the textual and thematic similarity of constitutions), a new domain comparison tool (to expand the scope of topics tracked in a corpus), and a recently released segments-as-topics tool (to formulate new topics and track them in a corpus), detailed below.
The research team developed and tested the tools as part of its own research to analyze constitutional content and expand the topics it tracks in constitutions. The team is now making the tools publicly available as open-source tools for others to start using as well.
Constitution Comparison Tool
The new constitution comparison tool helps researchers systematically assess the similarity of constitutions in two ways. First, it assesses whether constitutions show textual alignment—meaning, drafters use the same language across the constitutions. Second, it assesses whether constitutions show thematic alignment—meaning, the constitutions include similar topics regardless of wording.
This allows researchers to assess, for example: How much does a draft constitution change throughout the drafting process? How much of a new constitution is copied from the one it replaced? Does a constitution borrow language from other constitutions in the region? Do constitutions from a certain region or certain time period prioritize the same issues? Which constitutions are outliers in a set of constitutions?
Launching today, the constitution comparison tool is available in a public GitHub repository that includes instructions, applications, and sample data to facilitate others’ use of the tool. A video tutorial also walks users through how to use the tool in their own analysis.
Domain Comparison Tool
The new domain comparison tool helps researchers expand and improve the topics they track in texts. It does so by drawing topics from multiple ontologies—potentially representing multiple domains—and identifying these topics in the corpus under study.
This allows researchers to assess, for example: What is the conceptual overlap between two ontologies? Which parts of a corpus include topics that are not captured by its current ontology? Which topics in new ontologies could capture the meaning of that text? Do topics in new ontologies capture more nuances of an existing topic? Should topics from other domains be included in an ontology to capture the full content of a corpus?
Launching today, the domain comparison tool is available in a public GitHub repository that includes instructions, applications, and sample data to facilitate others’ use of the tool.
Segments-As-Topics Tool
The team’s segments-as-topics tool allows users to assess the best formulation of a potential new topic for a specific corpus, identify that topic in the text, and produce an exhaustive set of segments from the corpus that reference that topic. As the name of the tool implies, it uses segments from the corpus itself to represent topics that researchers seek to find in that corpus, producing more accurate automated matching and requiring human intervention at just a few key points.
Launched last fall, the segments-as-topics tool is also available in a public GitHub repository that includes instructions, applications, and sample data to facilitate others’ use of the tool. A video tutorial walks users through how to use the tool in their own analysis—whether in the constitutional domain or beyond.
Tools Training
These tools are part of a suite of open-source tools the program is developing for public use. The aim is to make these machine-assisted research methods accessible for a wide range of scholars and research applications across constitutional and comparative law.
The research team will host a training on these tools and the soon-to-be-released Sartori.Network concept repository. Register for the Concept Analysis Tools Training, taking place online on February 12.
The training will demo the tools, demonstrate varied use cases, and convey their relevance to scholars and practitioners engaged in research on a range of legal and constitutional issues. The training will also allow participants to discuss best practices related to ontologies, build working relationships, and provide a foundation for ongoing collaboration on concept integration.
Register to join the training on February 12 to learn how to use the tools and concept repository in your work!
—–
Funding acknowledgement: The Concept Integration in Comparative Law program is supported by National Science Foundation Award No. 2315189. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation (NSF). The research team deeply appreciates NSF’s Accountable Institutions and Behavior program and Human Networks and Data Science program for this support.
