ResearchGate Paper Download Tracker

WHAT IT DOES This script scans ResearchGate paper pages for download counts, reads, citations, and recommendations, and displays a floating panel showing a summary of paper statistics. It extracts download and read counts from the page, calculates total reads, downloads, and citation metrics. The panel displays a summary with color-coded indicators for engagement levels: green for high engagement, yellow for medium, red for low. A one-click export function saves paper statistics as CSV or JSON. All paper statistics snapshots are stored per domain locally for tracking over time. WHERE IT RUNS Works on ResearchGate paper pages (researchgate.net/publication/[id]). The script detects paper statistics using ResearchGate's page structure, including download counts, read counts, citation counts, recommendation counts, ResearchGate score, journal name, and publication date. HOW TO USE The panel appears automatically when you visit a ResearchGate paper page. It detects and displays the paper title, journal, publication date, and statistics grid showing downloads, reads, citations, and recommendations. The overall engagement level is color-coded based on total engagement (downloads + reads + citations). The ResearchGate score is shown when available. Click "Refresh Data" to rescan for updated statistics. Export paper statistics as CSV for spreadsheet analysis or as JSON for programmatic use. Use "Clear History" to reset stored snapshots for the current domain. The panel automatically saves a snapshot each time you visit, building a history of paper statistics over time. TECHNICAL MECHANISM: RESEARCHGATE PAPER STATISTICS EXTRACTION WITH ENGAGEMENT COLOR-CODING The script uses a comprehensive detection strategy. First, it extracts the paper title from h1 or title elements. Download counts are extracted using patterns like "[number] downloads" and "Full-text downloads: [number]." Read counts are extracted from patterns like "[number] reads" and "Readers: [number]." Citations are extracted from patterns like "[number] citations" and "Cited: [number]." Recommendations are extracted from patterns like "[number] recommendations" and "Recommended: [number]." The ResearchGate score is extracted from patterns like "RG Score: [number]" and "Score: [number]." The publication date is extracted from patterns like "Published: [date]" and "Publication date: [date]." The journal name is extracted from patterns like "Journal: [name]" and "Published in [name]." The script also scans elements with class names containing "stats," "stat," "metric," or "counter" to find additional statistic values. The overall engagement level is calculated as the sum of downloads, reads, and citations (excluding recommendations), and color-coded using thresholds: ≥500 = high (green), ≥100 = medium (yellow),