Countries
Institutions
Fields
Rankings
Compare

Search institutions, scholars, countries and publications

Sankore

Named after the University of Sankore in Timbuktu — one of the world’s oldest centres of learning. Every figure here is computed from work-level records in our own graph, never copied from a source’s summary count.

Built from

OpenAlex — works, authorships, subject classification
ROR — institutional identity
ORCID — researcher identity
Crossref — publisher confirmation

Browse

Countries
Institutions
Fields
Rankings
What we cover

Coverage is partial and says so. A country appears with its research output only once its works have been swept; until then the portal shows whatever the graph happens to hold for it and labels it as exactly that.

A Comparative Evaluation of Probabilistic and Transformer-Based Topic Models Across Diverse and Multilingual Text Corpora

2025Neural Processing LettersJournalOpen access

10.1007/s11063-025-11820-3 Full text3 citations

Authors

7 authors across 5 institutions in 2 countries.

Micheal Olalekan AjinajaFirst authorCorresponding
- The Federal Polytechnic, Ado-Ekiti (NG)
Johnson Tunde Fakoya
- Federal University of Agriculture, Abeokuta (NG)
Yetunde Esther Ogunwale
- Nigerian Geological Survey Agency (NG)
John Kolawole Omoniyi
Michael Adeniyi Ibiyomi
- The Federal Polytechnic, Ado-Ekiti (NG)
Akeem Adekunle Abiona
- The Federal Polytechnic, Ado-Ekiti (NG)
Damilola AkinolaSenior author

Fields

Assignment is probabilistic — a work belongs to several fields in proportions.

Computational and Text Analysis Methodslevel 3 · 0.42
Topic Modelinglevel 3 · 0.38
Sentiment Analysis and Opinion Mininglevel 3 · 0.04

Sources

Several sources describing one work is the point — it means the record rests on more than one authority.

openalexseen September 9, 2026
2025 · Neural Processing Letters · 3 citations
https://openalex.org/W4417435160

A Comparative Evaluation of Probabilistic and Transformer-Based Topic Models Across Diverse and Multilingual Text Corpora · Sankore