Comprehensive Analysis of Freebase and Dataset Creation for Robust Evaluation of Knowledge Graph Link Prediction Models

Freebase is amongst the largest public cross-domain knowledge graphs. It possesses three main data modeling idiosyncrasies. It has a strong type system; its properties are purposefully represented in reverse pairs; and it uses mediator objects to represent multiary relationships. These design choices are important in modeling the real-world. But they also pose nontrivial challenges in research of embedding models for knowledge graph completion, especially when models are developed and evaluated agnostically of these idiosyncrasies. This paper lays out a comprehensive analysis of the challenges associated with the idiosyncrasies of Freebase and measures their impact on knowledge graph link prediction. The results fill an important gap in our understanding of embedding models for link prediction as such models were never evaluated using a proper full-scale Freebase dataset. The paper also makes available several variants of the Freebase dataset by inclusion and exclusion of the data modeling idiosyncrasies. It fills an important gap in dataset availability too as this is the first-ever publicly available full-scale Freebase dataset that has gone through proper preparation.