OnT (Language Models as Ontology Encoder) - Complete Datasets

Complete dataset collection for OnT (Language Models as Ontology Encoder), a language model-based framework for ontology embeddings. This collection enables effective representation of concepts as points in hyperbolic space and axioms as hierarchical relationships between concepts. Built upon HierarchyTransformer, OnT provides enhanced capabilities through specialized embedding techniques. This collection includes 6 dataset files: ANATOMY_predict.zip: Training/validation for ANATOMY ontology ANATOMY_inference.zip: Inference/testing for ANATOMY ontology GALEN_predict.zip: Training/validation for GALEN ontology GALEN_inference.zip: Inference/testing for GALEN ontology GO_predict.zip: Training/validation for Gene Ontology GO_inference.zip: Inference/testing for Gene Ontology Each dataset contains preprocessed data for ontological reasoning tasks. Pre-trained models are available on Hugging Face: https://huggingface.co/collections/Hui97/ontology-transformer-68e8fdea10cba273bfdc687c