Cargando…

Learning-assisted theorem proving with millions of lemmas()

Large formal mathematical libraries consist of millions of atomic inference steps that give rise to a corresponding number of proved statements (lemmas). Analogously to the informal mathematical practice, only a tiny fraction of such statements is named and re-used in later proofs by formal mathemat...

Descripción completa

Detalles Bibliográficos
Autores principales: Kaliszyk, Cezary, Urban, Josef
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Elsevier Limited 2015
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4599631/
https://www.ncbi.nlm.nih.gov/pubmed/26525678
http://dx.doi.org/10.1016/j.jsc.2014.09.032
_version_ 1782394289562583040
author Kaliszyk, Cezary
Urban, Josef
author_facet Kaliszyk, Cezary
Urban, Josef
author_sort Kaliszyk, Cezary
collection PubMed
description Large formal mathematical libraries consist of millions of atomic inference steps that give rise to a corresponding number of proved statements (lemmas). Analogously to the informal mathematical practice, only a tiny fraction of such statements is named and re-used in later proofs by formal mathematicians. In this work, we suggest and implement criteria defining the estimated usefulness of the HOL Light lemmas for proving further theorems. We use these criteria to mine the large inference graph of the lemmas in the HOL Light and Flyspeck libraries, adding up to millions of the best lemmas to the pool of statements that can be re-used in later proofs. We show that in combination with learning-based relevance filtering, such methods significantly strengthen automated theorem proving of new conjectures over large formal mathematical libraries such as Flyspeck.
format Online
Article
Text
id pubmed-4599631
institution National Center for Biotechnology Information
language English
publishDate 2015
publisher Elsevier Limited
record_format MEDLINE/PubMed
spelling pubmed-45996312015-10-29 Learning-assisted theorem proving with millions of lemmas() Kaliszyk, Cezary Urban, Josef J Symb Comput Article Large formal mathematical libraries consist of millions of atomic inference steps that give rise to a corresponding number of proved statements (lemmas). Analogously to the informal mathematical practice, only a tiny fraction of such statements is named and re-used in later proofs by formal mathematicians. In this work, we suggest and implement criteria defining the estimated usefulness of the HOL Light lemmas for proving further theorems. We use these criteria to mine the large inference graph of the lemmas in the HOL Light and Flyspeck libraries, adding up to millions of the best lemmas to the pool of statements that can be re-used in later proofs. We show that in combination with learning-based relevance filtering, such methods significantly strengthen automated theorem proving of new conjectures over large formal mathematical libraries such as Flyspeck. Elsevier Limited 2015-07 /pmc/articles/PMC4599631/ /pubmed/26525678 http://dx.doi.org/10.1016/j.jsc.2014.09.032 Text en © 2014 The Authors http://creativecommons.org/licenses/by/3.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).
spellingShingle Article
Kaliszyk, Cezary
Urban, Josef
Learning-assisted theorem proving with millions of lemmas()
title Learning-assisted theorem proving with millions of lemmas()
title_full Learning-assisted theorem proving with millions of lemmas()
title_fullStr Learning-assisted theorem proving with millions of lemmas()
title_full_unstemmed Learning-assisted theorem proving with millions of lemmas()
title_short Learning-assisted theorem proving with millions of lemmas()
title_sort learning-assisted theorem proving with millions of lemmas()
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4599631/
https://www.ncbi.nlm.nih.gov/pubmed/26525678
http://dx.doi.org/10.1016/j.jsc.2014.09.032
work_keys_str_mv AT kaliszykcezary learningassistedtheoremprovingwithmillionsoflemmas
AT urbanjosef learningassistedtheoremprovingwithmillionsoflemmas