How to Extract Data from Scopus, WoS, and PubMed for Bibliometric Analysis

• ⏱️ 12 min read • 🗄️ Bibliometrics • Data Extraction

Did you know that 80% of bibliometric analysis failures don’t happen during mapping — they happen during data extraction? A wrong file format, incomplete metadata, or violated export limits can make VOSviewer and Biblioshiny reject your data — or worse, produce misleading maps. The principle is simple: garbage in, garbage out. This article guides you through extracting data from the three giant databases — Scopus, Web of Science (WoS), and PubMed — in formats compatible with popular bibliometric tools. If you don’t want to spend hours managing export limits and format conversions, the Corvexum team is ready to prepare a clean, analysis-ready dataset for you.

Step 0: Build a Solid Search String

Before touching the export button, make sure your query is mature. Use Boolean operators (AND, OR, NOT), quotation marks for exact phrases, and asterisks for truncation. Example: ("electric vehicle*" OR "ev adoption") AND ("consumer behavior" OR "purchase intention"). For PubMed, leverage MeSH terms for greater precision. Record and save your final search string — it is mandatory to report in your methodology section and reviewers will ask for it.

Extracting from Scopus (2,000 per Export Limit)

Scopus limits a single export to a maximum of 2,000 documents. If your results total 5,600 documents, you need 3 exports (1-2000, 2001-4000, 4001-5600). The steps:

  • Run the search, then select all results and click Export.
  • Choose the CSV format — the friendliest for Bibliometrix/Biblioshiny and VOSviewer.
  • Tick all metadata blocks: Citation information, Bibliographical information, Abstract & Keywords, and References. Skipping References means losing co-citation and bibliographic coupling capabilities.
  • Repeat for the next range, then save with sequential names (scopus_part1.csv, etc.).

Extracting from Web of Science (1,000 per Export Limit)

WoS Core Collection limits exports to 1,000 records per download. Use the Export → Excel / Plain Text / Tab-delimited menu. For bibliometric analysis, select Full Record and Cited References content. Both Plain Text (.txt) and Tab-delimited (.txt/.dat) formats are recognized by bibliometrix’s convert2df function. Don’t forget: WoS stores cited references most neatly of the three, making it superior for co-citation analysis and historiographs.

Extracting from PubMed (Free, but With Caveats)

PubMed is heaven for health and biomedical fields, and completely free. Use the Save / Send to → File button, then choose the PubMed or MEDLINE format that is directly compatible with bibliometrix. Download limits reach 10,000 records per session. But note two major limitations: PubMed does not provide cited references and has no built-in citation metrics. Solutions: combine with Europe PMC or OpenAlex for citation data, or use the NCBI E-utilities API for programmatic large-scale extraction.

Merging & Cleaning the Dataset

Once all files are collected, it’s consolidation time. In R, bibliometrix’s convert2df() function can merge Scopus/WoS files while performing automatic deduplication. Best steps:

  • Merge files per database first, then align column structures.
  • Remove duplicates based on DOI or title (tools: bibliometrix, or Rayyan for manual checks).
  • Check completeness of key fields: Authors, Title, Year, Source, Keywords, References. Empty fields will disable certain analysis menus.
  • Save the final dataset as a .RData object or master CSV — this is your project’s most valuable asset.
💡 Pro Documentation Tip: Create an extraction log table containing: database name, download date, search string, record count, and file format. This small table saves you when a reviewer asks “when and how was the data retrieved?” — and makes your study fully reproducible.

🎯 Which Database Should You Choose?

  1. Engineering, social sciences, multidisciplinary? → Scopus (widest coverage, 80+ million records).
  2. Need strong co-citation & historiograph? → WoS (neatest cited references).
  3. Medical/biomedical field, zero budget? → PubMed + Europe PMC/OpenAlex.
  4. Q1-class bibliometric study? → Use at least two databases and report the comparison — a major plus in reviewers’ eyes.

🚀 Don’t Want to Deal With Export Limits & Format Conversions?

Data extraction looks trivial, but a small mistake at this stage corrupts the entire analysis downstream — and fixing it means redoing all the work from scratch. Your time is too valuable to spend downloading files part 1 through part 5 while anxiously fearing omissions. The Corvexum team is ready to prepare a perfect data foundation:

  • ✅ Boolean search string design + validation with a research librarian
  • ✅ Complete Scopus/WoS/PubMed extraction including all export-limit parts
  • ✅ Deduplication, metadata cleaning, and master dataset consolidation
  • ✅ Ready-to-use files for VOSviewer, Biblioshiny, or CiteSpace + extraction log for methodology
💬 Free Consultation via WhatsApp
📧 Send Your Research Topic to Admin

Or email: corvexum@gmail.com | Visit: corvexum.com/

Crafted by the Corvexum bibliometrics team. Because great analysis always stands on correctly extracted data.


Tinggalkan Balasan

Alamat email Anda tidak akan dipublikasikan. Ruas yang wajib ditandai *

More posts

/* ============================================ CORVEXUM UNIVERSAL POST JS v2.2 ES5 + TANPA KARAKTER KHUSUS ( kurang-dari dan ampersand ) CTA Tracking + Smooth Scroll + PRISMA Animation ============================================ */ (function () { 'use strict'; function onReady(fn) { if (document.readyState !== 'loading') { fn(); } else { document.addEventListener('DOMContentLoaded', fn); } } onReady(function () { /* 1. CTA Click Tracking (siap untuk GA4/GTM) */ var btns = document.querySelectorAll('.cvx-btn'); var i; for (i = 0; i !== btns.length; i = i + 1) { (function (btn) { btn.addEventListener('click', function () { if (window.console) { if (window.console.log) { window.console.log('CTA Clicked: ' + btn.href); } } }); }(btns[i])); } /* 2. Smooth Scroll untuk anchor internal */ var anchors = document.querySelectorAll('a[href^="#"]'); var j; for (j = 0; j !== anchors.length; j = j + 1) { (function (link) { link.addEventListener('click', function (e) { var href = link.getAttribute('href'); var target = null; if (href) { if (href.length !== 1) { target = document.querySelector(href); } } if (target) { e.preventDefault(); target.scrollIntoView({ behavior: 'smooth', block: 'start' }); } }); }(anchors[j])); } /* 3. Animasi PRISMA Flow Diagram saat scroll into view */ var flows = document.querySelectorAll('.cvx-prisma-flow'); var m; var n; if (flows.length !== 0) { if (typeof window.IntersectionObserver === 'function') { var observer = new window.IntersectionObserver(function (entries) { for (n = 0; n !== entries.length; n = n + 1) { if (entries[n].isIntersecting) { entries[n].target.style.opacity = '1'; entries[n].target.style.transform = 'translateY(0)'; observer.unobserve(entries[n].target); } } }, { threshold: 0.1 }); for (m = 0; m !== flows.length; m = m + 1) { flows[m].style.opacity = '0'; flows[m].style.transform = 'translateY(20px)'; flows[m].style.transition = 'opacity 0.6s ease, transform 0.6s ease'; observer.observe(flows[m]); } } } /* 4. Auto-update tahun di footer (opsional) */ var yearSpans = document.querySelectorAll('.cvx-year'); var currentYear = new Date().getFullYear(); var p; for (p = 0; p !== yearSpans.length; p = p + 1) { yearSpans[p].textContent = String(currentYear); } }); }());