Re-SearchTerms is an interactive web platform for exploring how research terms in open scholarship, metascience, and research practice are defined across different sources. Built in collaboration with FORRT, it is deployed as a static HTML/JS app on GitHub Pages with a Supabase backend for live community contributions.
The platform enables users to:
- Compare definitions across sources (FORRT Glossary, Wiktionary, IGI, community submissions)
- Examine word-level co-occurrence patterns across definitions of a term
- Explore definition-level semantic similarity networks using precomputed SBERT embeddings
- Analyse term-level relationships through clustering and co-occurrence graphs
- Contribute source-based definitions in English, German, and Chinese
- Annotate and source-check existing definitions
- Track community contributions via the public contributor dashboard
Live app: forrt.org/re-searchterms
Re-SearchTerms was originally developed as part of the doctoral dissertation of Anna Yi Leung, which examined conceptual variability and terminological ambiguity in open scholarship and psychological science.
In collaboration with FORRT, the current version represents an expanded and evolving platform that aims to:
- Expand the dataset in alignment with the development of the FORRT Glossary
- Introduce new analytical features, including definition-level and term-level network analysis
- Support multilingual community contributions and public contributor recognition
- Provide an open, browser-based interface requiring no installation
- Dataset browser — Search and filter all terms; click any term to view its definitions across sources and languages
- Word-Level Analysis — Word frequency charts and co-occurrence networks across definitions of a selected term
- Definition-Level Network — SBERT cosine similarity graph connecting definitions of a term; includes live community definitions
- Term-Level Analysis — Word co-occurrence network, clustering graph, and types–tokens–TTR analysis across all terms
- Contribution system — Submit source-based definitions with full citation metadata; annotate and source-check existing definitions
- Contributor dashboard — Public leaderboard, contributor cards, activity feed, language coverage, and hall of fame
re-searchterms/
│
├── index.html # Main app (structure only)
├── styles.css # All CSS styles
├── app.js # All application JavaScript
├── env-config.js # Runtime env config — NOT committed (gitignored)
│ # Auto-generated by GitHub Actions at deploy time
│
├── data/ # Precomputed static datasets
│ ├── definition_nodes.json # Legacy definition node data
│ ├── definition_edges.json # Precomputed SBERT cosine similarity edges
│ ├── word_level_data.json # Word-level token data per term
│ └── term_level_data.json # Term-level clustering and TTR data
│
├── forrt_logo_inverted.png # FORRT logo (white version for dark header)
├── forrt_researchterms_smart_links.csv # Smart link data for cross-view navigation
│
├── .github/
│ └── workflows/
│ └── deploy.yml # GitHub Actions workflow for Pages deployment
│
├── Archived for R Shiny/ # Original Shiny app (archived)
├── Data Preprocessing/ # Data pipeline scripts
│
├── README.md
└── .gitignore
The app uses Supabase for:
- Authentication — Email/password login for registered contributors
- Definitions — Community-submitted source-based definitions (
definitionstable) - Annotations — Source checks and conceptual annotations (
definition_annotationstable) - Profiles — Public contributor profiles (
profilestable) - Terms — Canonical term list (
termstable)
Key views used by the frontend:
| View | Purpose |
|---|---|
v_definitions_public |
Approved definitions with annotation counts |
v_term_coverage |
Term list with definition counts per language |
v_contributor_profiles |
Public contributor profile data |
v_leaderboard |
Aggregated contributor statistics |
The data/ folder contains precomputed static datasets exported from the data pipeline. These power the Word-Level, Definition-Level, and Term-Level analysis views without requiring a server.
Raw datasets are not included due to size and source restrictions. To reproduce:
- Access original sources: FORRT Glossary, Wiktionary, IGI Global InfoScipedia
- Run scripts in
Data Preprocessing/
We welcome three types of contributions via the app:
- Source-original definitions — definitions as written in the original source (e.g., scholarly paper, book, glossary, report)
- Source-translated definitions — translations provided by the source itself
- Contributor-translated definitions — your own translation of a source definition
To contribute, sign up via the app and apply for contributor access. All contributions are publicly recognised.
Project Lead Anna Yi Leung — Ludwig-Maximilians-University of Munich, Germany
Co-Project Lead Dr. Daniel Kristanto — Carl von Ossietzky Universität Oldenburg, Germany
The project is open to collaboration, and additional contributors will be acknowledged as the platform evolves.
The development and evaluation of Re-SearchTerms has been accepted for publication:
Leung, A. Y., Kristanto, D., & Schmalz, X. (2026, accepted). Re-SearchTerms: A Shiny app for exploring terminology variations in psychology and metascience. Meta-Psychology.
For a detailed description of the conceptual framework, methods, and development of the app, please refer to our accepted paper here: https://osf.io/preprints/osf/qsp7x_v3
The initial development of Re-SearchTerms was supported by the META-REP Priority Programme and the German Research Foundation (DFG).
We thank Dr. Xenia Schmalz for her contributions to the first edition of the app.
The current version represents an independent and expanded development. We thank Dr. Flavio Azevedo (Director of FORRT) and Dr. Lukas Wallrich (Co-Chair of FORRT) for their technical feedback and for supporting the integration of the project within FORRT.
This project is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license unless stated otherwise.