Causal Structure Learning Supervised by Large Language Model
Causal discovery from observational data is pivotal for deciphering complex relationships. Causal Structure Learning (CSL), which focuses on deriving causal Directed Acyclic Graphs (DAGs) from data, faces challenges due to vast DAG spaces and data sparsity. The integration of Large Language Models (...
Saved in:
Main Authors | , , , , |
---|---|
Format | Journal Article |
Language | English |
Published |
20.11.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Causal discovery from observational data is pivotal for deciphering complex
relationships. Causal Structure Learning (CSL), which focuses on deriving
causal Directed Acyclic Graphs (DAGs) from data, faces challenges due to vast
DAG spaces and data sparsity. The integration of Large Language Models (LLMs),
recognized for their causal reasoning capabilities, offers a promising
direction to enhance CSL by infusing it with knowledge-based causal inferences.
However, existing approaches utilizing LLMs for CSL have encountered issues,
including unreliable constraints from imperfect LLM inferences and the
computational intensity of full pairwise variable analyses. In response, we
introduce the Iterative LLM Supervised CSL (ILS-CSL) framework. ILS-CSL
innovatively integrates LLM-based causal inference with CSL in an iterative
process, refining the causal DAG using feedback from LLMs. This method not only
utilizes LLM resources more efficiently but also generates more robust and
high-quality structural constraints compared to previous methodologies. Our
comprehensive evaluation across eight real-world datasets demonstrates
ILS-CSL's superior performance, setting a new standard in CSL efficacy and
showcasing its potential to significantly advance the field of causal
discovery. The codes are available at
\url{https://github.com/tyMadara/ILS-CSL}. |
---|---|
AbstractList | Causal discovery from observational data is pivotal for deciphering complex
relationships. Causal Structure Learning (CSL), which focuses on deriving
causal Directed Acyclic Graphs (DAGs) from data, faces challenges due to vast
DAG spaces and data sparsity. The integration of Large Language Models (LLMs),
recognized for their causal reasoning capabilities, offers a promising
direction to enhance CSL by infusing it with knowledge-based causal inferences.
However, existing approaches utilizing LLMs for CSL have encountered issues,
including unreliable constraints from imperfect LLM inferences and the
computational intensity of full pairwise variable analyses. In response, we
introduce the Iterative LLM Supervised CSL (ILS-CSL) framework. ILS-CSL
innovatively integrates LLM-based causal inference with CSL in an iterative
process, refining the causal DAG using feedback from LLMs. This method not only
utilizes LLM resources more efficiently but also generates more robust and
high-quality structural constraints compared to previous methodologies. Our
comprehensive evaluation across eight real-world datasets demonstrates
ILS-CSL's superior performance, setting a new standard in CSL efficacy and
showcasing its potential to significantly advance the field of causal
discovery. The codes are available at
\url{https://github.com/tyMadara/ILS-CSL}. |
Author | Wang, Xiangyu Lyu, Derui Ban, Taiyu Chen, Huanhuan Chen, Lyuzhou |
Author_xml | – sequence: 1 givenname: Taiyu surname: Ban fullname: Ban, Taiyu – sequence: 2 givenname: Lyuzhou surname: Chen fullname: Chen, Lyuzhou – sequence: 3 givenname: Derui surname: Lyu fullname: Lyu, Derui – sequence: 4 givenname: Xiangyu surname: Wang fullname: Wang, Xiangyu – sequence: 5 givenname: Huanhuan surname: Chen fullname: Chen, Huanhuan |
BackLink | https://doi.org/10.48550/arXiv.2311.11689$$DView paper in arXiv |
BookMark | eNotj8tOwzAURL2ABRQ-gBX-gQTfuH5JbFDESwpiEfbRTXwdRQpu5dQV_XtK6WpGGulozjW7iJtIjN2BKNdWKfGA6Wfal5UEKAG0dVfssca84MzbXcrDLifiDWGKUxx5m7eU9tNCnvcH3mAajyPGMeOxfGw8zTfsMuC80O05V6x9ef6q34rm8_W9fmoK1MYVWpDxRE45D6CIDOLaay0GZSAEayrbO2uCQu0AK-WscIOkoGTvB1JWrtj9P_X0vtum6RvTofuz6E4W8hcRiUO8 |
ContentType | Journal Article |
Copyright | http://arxiv.org/licenses/nonexclusive-distrib/1.0 |
Copyright_xml | – notice: http://arxiv.org/licenses/nonexclusive-distrib/1.0 |
DBID | AKY GOX |
DOI | 10.48550/arxiv.2311.11689 |
DatabaseName | arXiv Computer Science arXiv.org |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: GOX name: arXiv.org url: http://arxiv.org/find sourceTypes: Open Access Repository |
DeliveryMethod | fulltext_linktorsrc |
ExternalDocumentID | 2311_11689 |
GroupedDBID | AKY GOX |
ID | FETCH-LOGICAL-a679-60e7dee959d115ee7aa4d660c571ff8728b987f5a691a259809c3ef53bdce583 |
IEDL.DBID | GOX |
IngestDate | Mon Jan 08 05:48:21 EST 2024 |
IsDoiOpenAccess | true |
IsOpenAccess | true |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-a679-60e7dee959d115ee7aa4d660c571ff8728b987f5a691a259809c3ef53bdce583 |
OpenAccessLink | https://arxiv.org/abs/2311.11689 |
ParticipantIDs | arxiv_primary_2311_11689 |
PublicationCentury | 2000 |
PublicationDate | 2023-11-20 |
PublicationDateYYYYMMDD | 2023-11-20 |
PublicationDate_xml | – month: 11 year: 2023 text: 2023-11-20 day: 20 |
PublicationDecade | 2020 |
PublicationYear | 2023 |
Score | 1.897063 |
SecondaryResourceType | preprint |
Snippet | Causal discovery from observational data is pivotal for deciphering complex
relationships. Causal Structure Learning (CSL), which focuses on deriving
causal... |
SourceID | arxiv |
SourceType | Open Access Repository |
SubjectTerms | Computer Science - Artificial Intelligence |
Title | Causal Structure Learning Supervised by Large Language Model |
URI | https://arxiv.org/abs/2311.11689 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwdV1LT8MwDLa2nbggEKDxVA5cK9ouSROJC5oYE-JxKEi9TUntICSEpo1O8O9x2iK4cMnBycVOIvuLnc8A51hjSIPPEu08JjLwMTYGGbgW0tQaM_ZIMaN7_6Dnz_K2UtUAxM9fGLf6fN10_MB-fcHBR8aXWhs7hGGex5Ktm8eqS062VFz9-t91HGO2oj9OYrYD2310J6667diFAb3vweXUNWuWli1Va7Mi0ZOavoiyWca7uiYU_kvcxapsHrsXRBHblL3tQzm7fprOk75pQeJ0YROdUoFEVlnkWIuocE6i1mmtiiywJXLjGeUH5bTNHEMPk9p6QkFNPNakzOQARgz7aQzCKROcDiangFIG7RkpBS_ROqTCkj-EcavpYtnRUiyiERatEY7-nzqGrdgwPf6my9MTGLHmdMpu9cOftbb9Bqh_eNI |
link.rule.ids | 228,230,786,891 |
linkProvider | Cornell University |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Causal+Structure+Learning+Supervised+by+Large+Language+Model&rft.au=Ban%2C+Taiyu&rft.au=Chen%2C+Lyuzhou&rft.au=Lyu%2C+Derui&rft.au=Wang%2C+Xiangyu&rft.date=2023-11-20&rft_id=info:doi/10.48550%2Farxiv.2311.11689&rft.externalDocID=2311_11689 |