Generalized Reasoning With Graph Neural Networks by Relational Bayesian Network Encodings

Raffaele Pojer, Andrea Passerini, Manfred Jaeger
Proceedings of the Second Learning on Graphs Conference, PMLR 231:16:1-16:12, 2024.

Abstract

Graph neural networks (GNNs) and statistical relational learning are two different approaches to learning with graph data. The former can provide highly accurate models for specific tasks when sufficient training data is available, whereas the latter supports a wider range of reasoning types, and can incorporate manual specifications of interpretable domain knowledge. In this paper we present a method to embed GNNs in a statistical relational learning framework, such that the predictive model represented by the GNN becomes part of a full generative model. This model then supports a wide range of queries, including general conditional probability queries, and computing most probable configurations of unobserved node attributes or edges. In particular, we demonstrate how this latter type of queries can be used to obtain model-level explanations of a GNN in a flexible and interactive manner.

Cite this Paper

BibTeX
@InProceedings{pmlr-v231-pojer24a, title = {Generalized Reasoning With Graph Neural Networks by Relational Bayesian Network Encodings}, author = {Pojer, Raffaele and Passerini, Andrea and Jaeger, Manfred}, booktitle = {Proceedings of the Second Learning on Graphs Conference}, pages = {16:1--16:12}, year = {2024}, editor = {Villar, Soledad and Chamberlain, Benjamin}, volume = {231}, series = {Proceedings of Machine Learning Research}, month = {27--30 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v231/pojer24a/pojer24a.pdf}, url = {https://proceedings.mlr.press/v231/pojer24a.html}, abstract = {Graph neural networks (GNNs) and statistical relational learning are two different approaches to learning with graph data. The former can provide highly accurate models for specific tasks when sufficient training data is available, whereas the latter supports a wider range of reasoning types, and can incorporate manual specifications of interpretable domain knowledge. In this paper we present a method to embed GNNs in a statistical relational learning framework, such that the predictive model represented by the GNN becomes part of a full generative model. This model then supports a wide range of queries, including general conditional probability queries, and computing most probable configurations of unobserved node attributes or edges. In particular, we demonstrate how this latter type of queries can be used to obtain model-level explanations of a GNN in a flexible and interactive manner.} }
Endnote
%0 Conference Paper %T Generalized Reasoning With Graph Neural Networks by Relational Bayesian Network Encodings %A Raffaele Pojer %A Andrea Passerini %A Manfred Jaeger %B Proceedings of the Second Learning on Graphs Conference %C Proceedings of Machine Learning Research %D 2024 %E Soledad Villar %E Benjamin Chamberlain %F pmlr-v231-pojer24a %I PMLR %P 16:1--16:12 %U https://proceedings.mlr.press/v231/pojer24a.html %V 231 %X Graph neural networks (GNNs) and statistical relational learning are two different approaches to learning with graph data. The former can provide highly accurate models for specific tasks when sufficient training data is available, whereas the latter supports a wider range of reasoning types, and can incorporate manual specifications of interpretable domain knowledge. In this paper we present a method to embed GNNs in a statistical relational learning framework, such that the predictive model represented by the GNN becomes part of a full generative model. This model then supports a wide range of queries, including general conditional probability queries, and computing most probable configurations of unobserved node attributes or edges. In particular, we demonstrate how this latter type of queries can be used to obtain model-level explanations of a GNN in a flexible and interactive manner.
APA
Pojer, R., Passerini, A. & Jaeger, M.. (2024). Generalized Reasoning With Graph Neural Networks by Relational Bayesian Network Encodings. Proceedings of the Second Learning on Graphs Conference, in Proceedings of Machine Learning Research 231:16:1-16:12 Available from https://proceedings.mlr.press/v231/pojer24a.html.

Related Material