Augment to Interpret: Unsupervised and Inherently Interpretable Graph Embeddings

Gregory Scafarto, Madalina Ciortan, Simon Tihon, Quentin Ferre
Proceedings of the 15th Asian Conference on Machine Learning, PMLR 222:1183-1198, 2024.

Abstract

Unsupervised learning allows us to leverage unlabelled data, which has become abundantly available, and to create embeddings that are usable on a variety of downstream tasks. However, the typical lack of interpretability of unsupervised representation learning has become a limiting factor with regard to recent transparent-AI regulations. In this paper, we study graph representation learning and we show that data augmentation that preserves semantics can be learned and used to produce interpretations. Our framework, which we named INGENIOUS, creates inherently interpretable embeddings and eliminates the need for costly additional post-hoc analysis. We also introduce additional metrics addressing the lack of formalism and metrics in the understudied area of unsupervised-representation-learning interpretability. Our results are supported by an experimental study applied to both graph-level and node-level tasks and show that interpretable embeddings provide state-of-the-art performance on subsequent downstream tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v222-scafarto24a, title = {{Augment to Interpret}: {U}nsupervised and Inherently Interpretable Graph Embeddings}, author = {Scafarto, Gregory and Ciortan, Madalina and Tihon, Simon and Ferre, Quentin}, booktitle = {Proceedings of the 15th Asian Conference on Machine Learning}, pages = {1183--1198}, year = {2024}, editor = {Yanıkoğlu, Berrin and Buntine, Wray}, volume = {222}, series = {Proceedings of Machine Learning Research}, month = {11--14 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v222/scafarto24a/scafarto24a.pdf}, url = {https://proceedings.mlr.press/v222/scafarto24a.html}, abstract = {Unsupervised learning allows us to leverage unlabelled data, which has become abundantly available, and to create embeddings that are usable on a variety of downstream tasks. However, the typical lack of interpretability of unsupervised representation learning has become a limiting factor with regard to recent transparent-AI regulations. In this paper, we study graph representation learning and we show that data augmentation that preserves semantics can be learned and used to produce interpretations. Our framework, which we named INGENIOUS, creates inherently interpretable embeddings and eliminates the need for costly additional post-hoc analysis. We also introduce additional metrics addressing the lack of formalism and metrics in the understudied area of unsupervised-representation-learning interpretability. Our results are supported by an experimental study applied to both graph-level and node-level tasks and show that interpretable embeddings provide state-of-the-art performance on subsequent downstream tasks.} }
Endnote
%0 Conference Paper %T Augment to Interpret: Unsupervised and Inherently Interpretable Graph Embeddings %A Gregory Scafarto %A Madalina Ciortan %A Simon Tihon %A Quentin Ferre %B Proceedings of the 15th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Berrin Yanıkoğlu %E Wray Buntine %F pmlr-v222-scafarto24a %I PMLR %P 1183--1198 %U https://proceedings.mlr.press/v222/scafarto24a.html %V 222 %X Unsupervised learning allows us to leverage unlabelled data, which has become abundantly available, and to create embeddings that are usable on a variety of downstream tasks. However, the typical lack of interpretability of unsupervised representation learning has become a limiting factor with regard to recent transparent-AI regulations. In this paper, we study graph representation learning and we show that data augmentation that preserves semantics can be learned and used to produce interpretations. Our framework, which we named INGENIOUS, creates inherently interpretable embeddings and eliminates the need for costly additional post-hoc analysis. We also introduce additional metrics addressing the lack of formalism and metrics in the understudied area of unsupervised-representation-learning interpretability. Our results are supported by an experimental study applied to both graph-level and node-level tasks and show that interpretable embeddings provide state-of-the-art performance on subsequent downstream tasks.
APA
Scafarto, G., Ciortan, M., Tihon, S. & Ferre, Q.. (2024). Augment to Interpret: Unsupervised and Inherently Interpretable Graph Embeddings. Proceedings of the 15th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 222:1183-1198 Available from https://proceedings.mlr.press/v222/scafarto24a.html.

Related Material