Datasetv0.1 · Early Access
ILM Dataset
The Insight Language Machine is a CC BY-SA 4.0 licensed multilingual lexical knowledge graph. Available on HuggingFace as a gated dataset for verified researchers and AI teams.
Coverage
67,875
Lemmas
117,151
Senses
17
Languages
5
Corpora
Languages
mlമലയാളംenEnglishhiहिन्दीarالعربيةtaதமிழ்knಕನ್ನಡteతెలుగుsaसंस्कृतbnবাংলাmrमराठीurاردوorଓଡ଼ିଆdeDeutschfrFrançaisruРусскийzh中文ja日本語
Files
lemmas_hf_final.jsonlAll lemmas with core fields67,875✅
senses_hf_final.jsonlAll senses with definitions117,151✅
ilm_dataset.jsonlMultilingual glosses (17 lang)67,875 lemmas🔄 70% covered
senses_multilingual.jsonlSense-level multilingual (15 lang)48,080 / 117,151🔄 41% covered
Access
ILM is hosted on HuggingFace as a gated private dataset. Request access for AI training, NLP research, or academic use. Commercial licensing inquiries welcome.