Ivana Balažević


I am a final year PhD student in Machine Learning at the School of Informatics, University of Edinburgh, supervised by Tim Hospedales and Ivan Titov. I am a member of the Centre for Doctoral Training in Data Science. The focus of my PhD research has been on learning representations of entities and relations from knowledge graph data. Inspired by research in human cognition suggesting that humans reason about the world in terms of objects and their relations, I am interested in learning and reasoning over structured representations of the world. My research aim is to 1) extract discrete objects from sensory data and model their relationships; and 2) understand how semantics can be encoded in the geometry of representations. During my PhD, I interned at the Samsung AI Center Cambridge and at Facebook AI Research (FAIR) with Sebastian Riedel.


News: In November this year, I will be joining DeepMind as a Research Scientist!


Background

Before starting my PhD, I received a BSc in Computer Science at the Faculty of Electrical Engineering and Computing, University of Zagreb, an MSc in Computer Science at the Technische Universität Berlin and an MSc in Data Science at the University of Edinburgh.

Publications

Interpreting Knowledge Graph Relation Representation from Word Embeddings
Carl Allen*, Ivana Balažević*, Timothy Hospedales, ICLR, 2021
[paper]

Multi-relational Poincaré Graph Embeddings
Ivana Balažević, Carl Allen, Timothy Hospedales, NeurIPS, 2019
[paper] [code]

What the Vec? Towards Probabilistically Grounded Embeddings
Carl Allen, Ivana Balažević, Timothy Hospedales, NeurIPS, 2019
[paper]

TuckER: Tensor Factorization for Knowledge Graph Completion
Ivana Balažević, Carl Allen, Timothy Hospedales, EMNLP, 2019 (oral)
[paper] [code] [slides][video]

Hypernetwork Knowledge Graph Embeddings
Ivana Balažević, Carl Allen, Timothy Hospedales, ICANN, 2019 (oral)
[paper] [code] [slides]

Preprints

Cutting Down on Prompts and Parameters: Simple Few-Shot Learning with Language Models
Robert L. Logan IV, Ivana Balažević, Eric Wallace, Fabio Petroni, Sameer Singh, Sebastian Riedel arXiv:2106.13353, 2021
[paper] [code]

Learning the Prediction Distribution for Semi-Supervised Learning with Normalising Flows
Ivana Balažević*, Carl Allen*, Timothy Hospedales, arXiv:2007.02745, 2020
[paper] [code]

A Probabilistic Framework for Discriminative and Neuro-Symbolic Semi-Supervised Learning
Carl Allen, Ivana Balažević, Timothy Hospedales, arXiv:2006.05896, 2020
[paper]

Workshops

Benchmark and Best Practices for Biomedical Knowledge Graph Embeddings
David Chang, Ivana Balažević, Carl Allen, Daniel Chawla, Cynthia Brandt, Andrew Taylor
ACL BioNLP Workshop
, 2020
[paper]

TuckER: Tensor Factorization for Knowledge Graph Completion
Ivana Balažević, Carl Allen, Timothy Hospedales, ICML AMTL Workshop, 2019
[paper]

ibalazevic
ibalazevic

Contact

Email: ivana.balazevic@ed.ac.uk
Office: Room 3.50, Informatics Forum, 10 Crichton Street, Edinburgh, EH8 9AB