🚀 OneCLick Keyword Retriever

OneCLick is a lightweight, open-source Streamlit application for enriching bibliometric datasets by retrieving missing keywords from the OpenAlex API.

It streamlines preprocessing for tools like VOSviewer, CiteSpace, and Bibliometrix, and outputs reproducible, analysis-ready files for downstream research.

---

✨ Features

• 📁 Upload Input:

  • Accepts .xlsx files with a column named DI (DOI).

• 🔄 DOI-Based Keyword Retrieval:

  • Sends DOI queries to OpenAlex and retrieves enriched metadata keywords.

• ⚠️ Error & Warning Handling:

  • Invalid DOIs, network timeouts, or rate-limit errors → Failed_DOIs.csv
  • Valid DOIs without keywords → flagged as [WARNING:NoKeywords] in NoKeyword_DOIs.csv

• 📊 Latency Logging:

  • Per-DOI request latency is tracked and summarized (median & IQR)
  • Full request log saved as Latency_Log.csv

• 📤 Outputs:

  • Enriched dataset → openalex_keywords.csv / .xlsx
  • Downloadable error & warning logs
  • Latency logs for reproducibility and benchmarking

---

🛠️ Installation

# Clone the repository
git clone https://github.com/meMeta-a11y/oneclick.git
cd oneclick

# Install dependencies
pip install -r requirements.txt

Run the Streamlit application:

streamlit run streamlit_app.py

---

📊 Example Datasets

We provide three benchmark datasets for reproducibility and testing:

Dataset	DOIs	Purpose
Pilot	37	Quick exploratory testing
Intermediate	100	Lightweight benchmarking
Validation	1,097	Large-scale robustness testing

All datasets are available in the example_data/ directory and are permanently archived on Zenodo.

---

⚡ Performance Benchmarking

Validation experiments demonstrate robustness and scalability across dataset sizes:

Dataset	DOIs	Error DOIs	No-Keyword DOIs	Success Rate	Median Latency per DOI (seconds, IQR)
Pilot	37	0	0	100.00%	0.30 (0.29 – 0.31)
Intermediate	100	0	26	74.00%	0.30 (0.29 – 0.32)
Validation	1,097	0	205	81.31%	0.30 (0.29 – 0.32)

🧪 Benchmarks were conducted on Streamlit Cloud (Aug 2025) using a Linux environment (4 GB RAM, 100 Mbps network). Performance may vary depending on network conditions and OpenAlex API rate limits.

---

🚀 Quick-Start: VOSviewer Workflow

Follow these steps to integrate OneCLick outputs with VOSviewer:

1. Open VOSviewer → Create → Map based on network data
2. Input: Choose your enriched openalex_keywords.csv
3. Mapping: Set OpenAlex_KW as the keyword field
4. Options: Select co-occurrence counting method (e.g., Full counting)
5. Generate: Visualize the keyword network

📚 Additional workflows are described in the manuscript (Section 2.12).

---

📁 Repository Structure

oneclick/
├── streamlit_app.py         # Main application
├── requirements.txt         # Dependencies
├── example_data/            # Example DOI datasets
│   ├── pilot_37.xlsx
│   ├── intermediate_100.xlsx
│   └── validation_1097.xlsx
├── outputs/                 # Example outputs & logs
│   ├── openalex_keywords.csv
│   ├── Failed_DOIs.csv
│   ├── NoKeyword_DOIs.csv
│   └── Latency_Log.csv
└── README.md               # Documentation

---

📚 Citation

If you use OneCLick in your research, please cite it using one of the following formats:

📘 APA

Wei, L. K. (2025). OneCLick: Streamlined metadata enrichment using machine-inferred keywords from OpenAlex. SoftwareX, 31, 102353.

📙 Chicago

Wei, Loo Keat. "OneCLick: Streamlined metadata enrichment using machine-inferred keywords from OpenAlex." SoftwareX 31 (2025): 102353.

📗 Harvard

Wei, L.K., 2025. OneCLick: Streamlined metadata enrichment using machine-inferred keywords from OpenAlex. SoftwareX, 31, p.102353.

📕 Vancouver

Wei LK. OneCLick: Streamlined metadata enrichment using machine-inferred keywords from OpenAlex. SoftwareX. 2025 Sep 1;31:102353.

---

🌐 About

• 🔗 Live Demo: metakey.streamlit.app
• 📦 Source Code: GitHub Repository

---

📜 License

This project is licensed under the Apache-2.0 License - see the LICENSE file for details.

---

💡 OneCLick is developed to accelerate bibliometric research workflows by automating metadata enrichment and preparing datasets for network analysis and visualization. Contributions, issues, and pull requests are always welcome!