Probabilistic Similarity Logic

Matthias Broecheler, Lilyana Mihalkova, Lise Getoor
Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, PMLR R8:73-82, 2010.

Abstract

Many machine learning applications require the ability to learn from and reason about noisy multi-relational data. To address this, several ef- fective representations have been developed that provide both a language for expressing the struc- tural regularities of a domain, and principled sup- port for probabilistic inference. In addition to these two aspects, however, many applications also involve a third aspect–the need to reason about similarities–which has not been directly supported in existing frameworks. This paper introduces probabilistic similarity logic (PSL), a general-purpose framework for joint reason- ing about similarity in relational domains that incorporates probabilistic reasoning about sim- ilarities and relational structure in a principled way. PSL can integrate any existing domain- specific similarity measures and also supports reasoning about similarities between sets of en- tities. We provide efficient inference and learn- ing techniques for PSL and demonstrate its ef- fectiveness both in common relational tasks and in settings that require reasoning about similarity.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR8-broecheler10a, title = {Probabilistic Similarity Logic}, author = {Broecheler, Matthias and Mihalkova, Lilyana and Getoor, Lise}, booktitle = {Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence}, pages = {73--82}, year = {2010}, editor = {Grünwald, Peter and Spirtes, Peter}, volume = {R8}, series = {Proceedings of Machine Learning Research}, month = {08--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r8/main/assets/broecheler10a/broecheler10a.pdf}, url = {https://proceedings.mlr.press/r8/broecheler10a.html}, abstract = {Many machine learning applications require the ability to learn from and reason about noisy multi-relational data. To address this, several ef- fective representations have been developed that provide both a language for expressing the struc- tural regularities of a domain, and principled sup- port for probabilistic inference. In addition to these two aspects, however, many applications also involve a third aspect–the need to reason about similarities–which has not been directly supported in existing frameworks. This paper introduces probabilistic similarity logic (PSL), a general-purpose framework for joint reason- ing about similarity in relational domains that incorporates probabilistic reasoning about sim- ilarities and relational structure in a principled way. PSL can integrate any existing domain- specific similarity measures and also supports reasoning about similarities between sets of en- tities. We provide efficient inference and learn- ing techniques for PSL and demonstrate its ef- fectiveness both in common relational tasks and in settings that require reasoning about similarity.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Probabilistic Similarity Logic %A Matthias Broecheler %A Lilyana Mihalkova %A Lise Getoor %B Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2010 %E Peter Grünwald %E Peter Spirtes %F pmlr-vR8-broecheler10a %I PMLR %P 73--82 %U https://proceedings.mlr.press/r8/broecheler10a.html %V R8 %X Many machine learning applications require the ability to learn from and reason about noisy multi-relational data. To address this, several ef- fective representations have been developed that provide both a language for expressing the struc- tural regularities of a domain, and principled sup- port for probabilistic inference. In addition to these two aspects, however, many applications also involve a third aspect–the need to reason about similarities–which has not been directly supported in existing frameworks. This paper introduces probabilistic similarity logic (PSL), a general-purpose framework for joint reason- ing about similarity in relational domains that incorporates probabilistic reasoning about sim- ilarities and relational structure in a principled way. PSL can integrate any existing domain- specific similarity measures and also supports reasoning about similarities between sets of en- tities. We provide efficient inference and learn- ing techniques for PSL and demonstrate its ef- fectiveness both in common relational tasks and in settings that require reasoning about similarity. %Z Reissued by PMLR on 04 October 2026.
APA
Broecheler, M., Mihalkova, L. & Getoor, L.. (2010). Probabilistic Similarity Logic. Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R8:73-82 Available from https://proceedings.mlr.press/r8/broecheler10a.html. Reissued by PMLR on 04 October 2026.

Related Material