Embed2Sym: scalable neuro-symbolic reasoning via clustered embeddings

Mostra el registre complet Registre parcial de l'ítem

• dc.contributor.author Aspis, Yaniv
• dc.contributor.author Broda, Krysia
• dc.contributor.author Lobo, Jorge
• dc.contributor.author Russo, Alessandra
• dc.date.accessioned 2023-04-05T06:22:37Z
• dc.date.available 2023-04-05T06:22:37Z
• dc.date.issued 2022
• dc.description Comunicació presentada a 19th International Conference on Principles of Knowledge Representation and Reasoning (KR2022), celebrat del 31 de juliol al 5 d'agost de 2022 a Haifa, Israel.
• dc.description.abstract Neuro-symbolic reasoning approaches proposed in recent years combine a neural perception component with a symbolic reasoning component to solve a downstream task. By doing so, these approaches can provide neural networks with symbolic reasoning capabilities, improve their interpretability and enable generalization beyond the training task. However, this often comes at the cost of poor training time, with potential scalability issues. In this paper, we propose a scalable neuro-symbolic approach, called Embed2Sym. We complement a two-stage (perception and reasoning) neural network architecture designed to solve a downstream task end-to-end with a symbolic optimisation method for extracting learned latent concepts. Specifically, the trained perception network generates clusters in embedding space that are identified and labelled using symbolic knowledge and a symbolic solver. With the latent concepts identified, a neuro-symbolic model is constructed by combining the perception network with the symbolic knowledge of the downstream task, resulting in a model that is interpretable and transferable. Our evaluation shows that Embed2Sym outperforms state-of-the-art neuro-symbolic systems on benchmark tasks in terms of training time by several orders of magnitude while providing similar if not better accuracy.
• dc.format.mimetype application/pdf
• dc.identifier.citation Aspis Y, Broda K, Lobo J, Russo A. Embed2Sym: scalable neuro-symbolic reasoning via clustered embeddings. In: Kern-Isberner G, Lakemeyer G, Meyer T, editors. Proceedings of the 19th International Conference on Principles of Knowledge Representation and Reasoning (KR2022); 2022 Jul 31-Aug 5; Haifa, Israel. [California]: IJCAI Organization; 2022. p. 421-31. DOI: 10.24963/kr.2022/44
• dc.identifier.doi http://dx.doi.org/10.24963/kr.2022/44
• dc.identifier.isbn 978-1-956792-01-0
• dc.identifier.issn 2334-1033
• dc.identifier.uri http://hdl.handle.net/10230/56421
• dc.language.iso eng
• dc.publisher International Joint Conferences on Artificial Intelligence Organization
• dc.relation.ispartof Kern-Isberner G, Lakemeyer G, Meyer T, editors. Proceedings of the 19th International Conference on Principles of Knowledge Representation and Reasoning (KR2022); 2022 Jul 31-Aug 5; Haifa, Israel. [California]: IJCAI Organization; 2022. p. 421-31.
• dc.rights © 2022 International Joint Conferences on Artificial Intelligence. This is the published version of a paper available at http://dx.doi.org/10.24963/kr.2022/44 that appeared in the Proceedings of the 19th International Conference on Principles of Knowledge Representation and Reasoning.
• dc.rights.accessRights info:eu-repo/semantics/openAccess
• dc.subject.keyword Neural-symbolic learning
• dc.subject.keyword Integrating symbolic and sub-symbolic approaches
• dc.subject.keyword Learning symbolic abstractions from unstructured data
• dc.subject.keyword Interplay between logic & neural and other learning paradigms
• dc.title Embed2Sym: scalable neuro-symbolic reasoning via clustered embeddings
• dc.type info:eu-repo/semantics/conferenceObject
• dc.type.version info:eu-repo/semantics/publishedVersion