MikeACedric/domain_classifier_demo

ORKG Core Domain Classifier — Metadata

Update Date: 2026-09-20

Research-domain labels produced by the ORKG Core Domain Classifier for documents in the CORE corpus. One row per document: the document id and the domain the classifier assigned to it.

The classifier that generates this dataset lives at https://gitlab.com/TIBHannover/orkg/nlp/experiments/core-domain-classifier.

This dataset is rebuilt by python -m app.run_hf_sync from that repository — new classifications are added, re-classified documents are overwritten in place, and each run is tagged with a version. Do not edit the Parquet files by hand.

📊 Current Stats — version v4 (2026-09-20)

Total Classified Records: 8,732,353

Classification Label	Count	Percentage (%)
Arts and Humanities	1,124,474	12.88%
Engineering	370,709	4.25%
Life Sciences	306,968	3.52%
Physical Sciences and Mathematics	154,592	1.77%
Social and Behavioral Sciences	2,136,312	24.46%
Aerospace Engineering	70,005	0.80%
Biomedical Engineering and Bioengineering	1,103,084	12.63%
Chemical Engineering	112,523	1.29%
Civil and Environmental Engineering	207,966	2.38%
Computer Engineering	224,589	2.57%
Computational Engineering	134,410	1.54%
Electrical Engineering and Information Technology	195,501	2.24%
Electrical and Computer Engineering	119,870	1.37%
Energy Process Engineering	45,581	0.52%
Engineering Science and Materials	171,741	1.97%
Materials Engineering	23,530	0.27%
Materials Science and Engineering	6,159	0.07%
Mechanical Engineering	30,712	0.35%
Nanoscience and Nano Technology	5,551	0.06%
Nuclear Engineering	210,187	2.41%
Operations Research, Systems Engineering and Industrial Engineering	85,787	0.98%
Production Technology	5,863	0.07%
Plastics Engineering	433	0.00%
American Studies	77,531	0.89%
Classics	6,702	0.08%
Educational Research	63,452	0.73%
English Language and Literature	12,841	0.15%
Religion	6,869	0.08%
History	61,029	0.70%
History of Art, Architecture and Archaelogy	46,901	0.54%
Literary Studies	4,405	0.05%
Musicology	1,626	0.02%
Philosophy	26,733	0.31%
Animal Sciences	14,016	0.16%
Agriculture, Forestry and Veterinary Medicine	125,851	1.44%
Biochemistry, Biophysics, and Structural Biology	26,136	0.30%
Bioinformatics	1,597	0.02%
Biotechnolgy	2	0.00%
Public Health	97,974	1.12%
Systems Biology	6,688	0.08%
Physiology	1,455	0.02%
Nutrition	1,664	0.02%
Nursing Pharmacology, Toxicology and Environmental Health	25,476	0.29%
Neuroscience and Neurobiology	35,270	0.40%
Microbiology	5,600	0.06%
Medicine	96,337	1.10%
Immunology and Infectious diseases	21,169	0.24%
Genetics and Genomics	2,034	0.02%
Forestry and Forest Sciences	12,303	0.14%
Entomology Food Science	13,302	0.15%
Ecology and Evolutionary Biology	35,307	0.40%
Cell and Developmental Biology	3,699	0.04%
Water Research	9,844	0.11%
Statistics and Probability	8,486	0.10%
Polymer Research	265	0.00%
Statistical Physics, Soft Matter, Biological Physics, Nonlinear Dynamics	290	0.00%
Physics	171,522	1.96%
Particles, Nuclei and Fields	936	0.01%
Optics, Quantum Optics and Physics of Atoms, Molecules and Plasmas	4,273	0.05%
Oceanography and Atmospheric Sciences and Meteorology	50,942	0.58%
Mineralogy, Petrology and Geochemistry	778	0.01%
Oceanography	1,148	0.01%
Mathematics	85,017	0.97%
Computer Sciences	198,701	2.28%
Earth Sciences	94,322	1.08%
Applied Mathematics	6,449	0.07%
Chemistry	18,998	0.22%
Chemical Solid State and Surface Research	227	0.00%
Astrophysics and Astronomy	44,353	0.51%
Urban Studies and Planning	29,657	0.34%
Sociology	468	0.01%
Social and Cultural Anthropology and Ethnology	13,311	0.15%
Social Sciences	98,622	1.13%
Political Science	10,755	0.12%
Psychology	7,787	0.09%
Public Affairs, Public Policy and Public Administration	44,019	0.50%
Linguistics	3,110	0.04%
Jurisprudence	3,012	0.03%
Economics	72,952	0.84%
Communication Sciences	31,718	0.36%
Agricultural and Resource Economics	10,521	0.12%
Language, Socieities and Cultures	27,647	0.32%
Comparative Literature	1,707	0.02%

Schema

Column	Description
id	string — CORE document id
predicted_domain	string — research domain assigned by the classifier
updated_at	string (YYYY-MM-DD) — date this row was last written

Usage

from datasets import load_dataset

dataset = load_dataset("MikeACedric/domain_classifier_demo", split="train")

# a specific release (every sync is tagged):
pinned = load_dataset("MikeACedric/domain_classifier_demo", split="train", revision="v4")

Version history

Version	Update Date	Rows	Added	Updated	Deleted
v4	2026-09-20	8,732,353	0	200,000	0
v3	2026-09-15	8,732,353	199,945	55	0
v2	2026-09-01	8,532,408	0	100,000	1,000
v1	2026-08-06	8,533,408	8,533,408	0	0