Metadata for Metacognition: Supporting Self-Regulation Capabilities in Soar
Proceedings of the AAAI Symposium Series · 06.11.2026 · 10 (1) · 133-137
DOI: 10.1609/aaaiss.v10i1.43400 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Symposium Series, online erschienen am 06.11.2026. Verfasst von Steven J. Jones und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Metacognition, the ability to represent, monitor, and modulate cognitive processing, is realized in the Soar cognitive architecture through the interaction of metadata and standard architecture mechanisms for reasoning, now applied to reasoning (metareasoning). To explore enabling the wider range of self-regulatory metacognitive capabilities reported in humans, we examine what metadata is required and the current support (or lack thereof) in Soar for this metadata. We propose an approach where metadata is produced by appraisal processes that evaluate agent state in the context of current tasking. We present an initial investigation, where we implement a limited set of appraisals that signal agent progress (or lack thereof) in a troubleshooting domain, facilitating agent strategy change when it is unable to make progress.
Zitieren
APA
Jones, S. J., Kirk, J. R., & Wray, R. E. (2026). Metadata for Metacognition: Supporting Self-Regulation Capabilities in Soar. Proceedings of the AAAI Symposium Series, 10(1), 133-137. https://doi.org/10.1609/aaaiss.v10i1.43400
Vancouver
Jones SJ, Kirk JR, Wray RE. Metadata for Metacognition: Supporting Self-Regulation Capabilities in Soar. Proceedings of the AAAI Symposium Series. 2026;10(1):133-137. doi:10.1609/aaaiss.v10i1.43400
BibTeX
@article{Jones2026,
author = {Jones, Steven J. and Kirk, James R. and Wray, Robert E.},
title = {Metadata for Metacognition: Supporting Self-Regulation Capabilities in Soar},
journal = {Proceedings of the AAAI Symposium Series},
year = {2026},
volume = {10},
number = {1},
pages = {133--137},
doi = {10.1609/aaaiss.v10i1.43400}
}
EUVIMED