Schema-aware Translating Embeddings model

We implemented a Torch Deep Learning model able to learn Knowledge Graph embeddings from knowledge base triples. We modified the original negative sampling procedure presented in the paper Translating Embeddings for Modeling Multi-relational Data by incorporating constraints derived from the Ontology schema in order to learn more effective and expressive representations.

Project link:

Nifty tech tag lists fromĀ Wouter Beeftink