Tutorial Material and Recording

:file_folder: Slides [PDF]

:pencil2: Jupyter notebook/Colab Hands-On Session

:pencil2: Check out additional KGE tutorial material

Abstract

Knowledge graph embeddings are supervised learning models that learn vector representations of nodes and edges of labeled, directed multigraphs. We describe their design rationale, and explain why they are receiving growing attention within the burgeoning graph representation learning community. We highlight their limitations, open research directions, and real-world applicative scenarios. Besides a theoretical overview, we also provide a hands-on session, where we show how to use such models in practice.

Goal

The goal of the tutorial is to provide answers to the following questions:

Outline

Slides available here

Target Audience

The target audience is the general ECAI audience who is interested in knowledge representation and reasoning, machine learning, and natural language processing.

That includes artificial intelligence scientists, engineers, and students familiar with neural networks fundamentals and eager to know insights of graph representation learning for knowledge graphs. Researchers from graph-based knowledge representation (e.g. Semantic Web, Linked Data) and NLP also qualify as target audience.

The tutorial is of interest for either academic research and industry practitioners.

Cite as

@misc{kge_tutorial,
  author       = {Luca Costabello and
                  Sumit Pai and
                  Nicholas McCarthy and
                  Adrianna Janik},
  title        = {Knowledge Graph Embeddings Tutorial: From Theory to Practice},
  month        = sep,
  year         = 2020,
  note         = {https://kge-tutorial-ecai2020.github.io/},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.4268208},
  url          = {https://doi.org/10.5281/zenodo.4268208}
}

Organizers

Luca Costabello is research scientist in Accenture Labs Dublin. His research interests span knowledge graphs applications, machine learning for graphs, and explainable AI.


Sumit Pai is a research engineer at Accenture Labs Dublin. His research interests include knowledge graphs, representational learning, computer vision and its applications. Sumit has also worked as an engineer (Computer Vision) at Robert Bosch, India. He has done his Masters in Neural Information Processing from University of Tübingen, Germany.


Nicholas McCarthy is a research scientist at Accenture Labs. His reseearch interests include computer vision, graph representation learning, data privacy and medical imaging. Prior to joining Accenture Labs Nicholas worked at the INSIGHT Research Center and the Complex and Adaptive Systems Laboratory in University College Dublin, Ireland.


Adrianna Janik is a research engineer at Accenture Labs Dublin. Her research interests are interpretability in machine learning, deep learning, and recently knowledge graphs. She has double Masters in Data Science with a minor in entrepreneurship from the European Institute of Innovation and Technology (EIT), at the University of Nice - Sophia Antipolis and at the Royal Institute of Technology, Stockholm. During studies, she did her thesis internship at the Montreal Institute for Learning Algorithms. She also has a Bachelors in Control Engineering and Robotics from the Wroclaw University of Technology and used to work as a software engineer at Nokia.