▶ Merge Web of Science + Scopus in Google Colab — nothing to install.
Open the notebook → run cell 1 → run cell 2 → pick your project.

A lightweight, terminal / Colab tool that merges Web of Science + Scopus exports into a single deduplicated dataset using BibexPy's Smart Merge algorithm (DOI-determinative deduplication).

It is the "just merge" companion to the full BibexPy app: same merge algorithm, no web UI, no API/ML enrichment — ideal for a quick run in Google Colab or a terminal, the way BibexPy v1 worked.

What it does: read raw WoS .txt + Scopus .csv → merge & deduplicate → write Merged.xlsx and a VOSviewer / biblioshiny-ready Merged_Vos.txt. What it does NOT do: enrichment, APIs, filtering, harmonization, reporting. For those, use the full BibexPy app.

---

Smart Merge in one line

Records are matched in stages: negative rules → DOI exact → PMID exact → Title+Year+Surname → Journal+Volume+Pages → borderline. DOI is determinative — two records whose normalized DOIs differ are never the same publication (no auto-merge, no false dedup). Field values are combined with fixed per-field source preferences (e.g. abstract/authors from Scopus, citations from Web of Science).

Uncertain pairs (title similarity 0.80–0.92) are kept separate and written to Borderline_Uncertain.xlsx for you to review manually.

---

▶ Run in Google Colab (recommended — no install)

1. Click the badge → the notebook opens in Colab.
2. Run cell 1 (setup) and cell 2 (merge) — pick your project from the menu.
3. Try the bundled Sample Project, or upload your own WoS .txt + Scopus .csv with the optional cell, then run cell 2.

Results are saved under Workspace/<project>/Analysis_<timestamp>/.

Use in a terminal

git clone https://github.com/bcankara/BibexPy-Lite.git
cd BibexPy-Lite
pip install -r requirements.txt
python merge.py

Project layout

Put your raw exports under Workspace/<Your Project>/Data/:

Workspace/
  My Project/
    Data/
      savedrecs.txt      # Web of Science plain-text export(s) — one or more
      scopus.csv         # Scopus CSV export(s) — one or more

Run python merge.py, pick your project from the numbered menu, and the results are written to Workspace/My Project/Analysis_<timestamp>/:

File	Description
Merged.xlsx	Final deduplicated dataset
Merged_Vos.txt	WoS-tagged text for VOSviewer / biblioshiny
Borderline_Uncertain.xlsx	Uncertain pairs kept separate (review)
Conflict_Log.xlsx	Field conflicts resolved during merge
Lost_WoS.xlsx / Lost_Scopus.xlsx	Records with no match
Statistics.xlsx	Summary counts

Use as a library

from bibexpy_lite import read_wos, read_scopus, smart_merge

wos = read_wos("Workspace/My Project/Data/savedrecs.txt")
scp = read_scopus("Workspace/My Project/Data/scopus.csv")
res = smart_merge(wos, scp)

res.merged.to_excel("merged.xlsx", index=False)
print(res.stats)            # counts: duplicates_removed, merged_count, ...
print(res.borderline)       # uncertain pairs

---

Relationship to the main BibexPy

The Smart Merge algorithm here (bibexpy_lite/smart_merge.py) is a vendored copy of the canonical implementation in BibexPy (apps/api/services/smart_merger.py). The algorithm is the single source of truth there; this repo keeps a copy so it can run with no web dependencies. They are kept in sync and produce identical merge results.

Citation

BibexPy-Lite uses the BibexPy Smart Merge algorithm. If you use it in your research, please cite:

Kara, B. C., Şahin, A., & Dirsehan, T. (2025). BibexPy: Harmonizing the bibliometric symphony of Scopus and Web of Science. SoftwareX, 30, 102098. https://doi.org/10.1016/j.softx.2025.102098

@article{bibexpy2025,
  title   = {BibexPy: Harmonizing the bibliometric symphony of {Scopus} and {Web of Science}},
  author  = {Kara, Burak Can and {\c{S}}ahin, Alperen and Dirsehan, Ta{\c{s}}k{\i}n},
  journal = {SoftwareX},
  volume  = {30},
  pages   = {102098},
  year    = {2025},
  doi     = {10.1016/j.softx.2025.102098}
}

License

GPL-3.0-or-later. See LICENSE.