TiAb Review Plugin: A Browser-Based Tool for AI-Assisted Title and Abstract Screening
Summary: arXiv:2604.08602v1 Announce Type: cross
Abstract
Background: Server-based screening tools impose subscription costs, while open-source alternatives require coding skills.
Objectives: We developed a browser extension that provides no-code, serverless artificial intelligence (AI)-assisted title and abstract screening and examined its functionality.
Methods
The TiAb Review Plugin is an open-source Chrome browser extension.
It is available at Chrome Web Store.
The plugin uses Google Sheets as a shared database, requiring no dedicated server and enabling multi-reviewer collaboration.
Users supply their own Gemini API key, which is stored locally and encrypted.
The tool offers three screening modes:
- Manual review
- Large language model (LLM) batch screening
- Machine learning (ML) active learning
For ML evaluation, we re-implemented the default ASReview active learning algorithm (TF-IDF with Naive Bayes) in TypeScript to enable in-browser execution.
We verified the equivalence against the original Python implementation using 10-fold cross-validation on six datasets.
For LLM evaluation, we compared 16 parameter configurations across two model families on a benchmark dataset.
Subsequently, we validated the optimal configuration (Gemini 3.0 Flash, low thinking budget, TopP=0.95) with a sensitivity-oriented prompt on five public datasets, containing between 1,038 to 5,628 records and a prevalence ranging from 0.5 to 2.0 percent.
Results
The TypeScript classifier produced top-100 rankings that were 100 percent identical to the original ASReview across all six datasets.
For LLM screening, we observed a recall rate ranging from 94 to 100 percent, with precision levels between 2 to 15 percent.
The Work Saved over Sampling at 95 percent recall (WSS@95) ranged from 48.7 to 87.3 percent, indicating significant efficiency in the screening process.
Conclusions
We developed a functional browser extension that integrates LLM screening and ML active learning into a no-code, serverless environment.
The TiAb Review Plugin is prepared for practical use in systematic review screening, offering a cost-effective and efficient solution for researchers.