KGvec2go - Knowledge Graph Embeddings as a Service

https://arxiv.org/pdf/2003.05809.pdf

• Source: http://kgvec2go.org/

Basic Idea

A web API for accessing and consuming graph embeddings in a light-weight fashion in downstream applications. Also, use a combination of multiple models which leads to better outcome.

Approach

• Walk generation of embeddings (duplicate free random walks)

• Web API to access the data models in a lightweight way. Bring powerful embedding models to devices with restriction in CPU and RAM.

Question

• What is its motivation and key contribution?

• How does the access work exactly? (computationally cheap, memory efficient, easy), how does this paper show all of these?