LegalRuleML standard approved by OASIS

News and Events
#1

Announcement here.

LegalRuleML is an XML representation of semantic encodings of legislation built on top of RuleML, which is used as a transfer protocol for production rule and logic systems. LegalRuleML adds features like references to source material, documentation of encoders’ identities, alternative interpretations, in-force dates, and other features particularly useful for statutes.

While it contains a lot of the same information as many executable languages, the LegalRuleML standard does not include a semantics, and so technically is not executable on its own. It’s function seems to be as a target for import and export functionalities in different tools, or as a storage language in non-executable applications, like dependency analysis.

Note that this is different from Akoma Ntoso, also known as LegalDocML, which is an XML standard for representing legal documents. So if you have a law, and you encode it in Prolog, the law could be represented in Akoma Ntoso, and the Prolog encoding could be represented in LegalRuleML.

Congrats to the LegalRuleML team!

#2

Hi Jason, I would love to understand the capabilities and differences in these XML standards. Is there a good overview, or could you do a Q&A session sometime?

#3

Hey, @heidir! Thanks for the question.

Off the top of my head, I don’t know of any good introductory-level documentation for either of them. I’m aware of the standards themselves, and I know there are some papers out there that talk about their development. AN is actually used in a number of real-world applications, particularly systems for managing legislative drafting life-cycles, but it’s used as a storage format for the legislation, so it is mostly invisible to the user. I’m not aware of LegalRuleML being deployed anywhere.

A Q&A with me wouldn’t get you far! :slight_smile:

If you find anything, please add it to our knowledge base!

#4

It does seem to me that lrml & AN do more than this though. The combination of lrml:PrescriptiveStatement and lrml:Obligation elements, with if-then logic elements seems to me to capture what drafters should be marking up in our texts. They try to do that for all sorts of law, and to capture all sorts of other info, to the point where it ends up extremely cumbersome and the work seems to be done in big chunks of preliminary metadata instead of inline markup in the natural language text. See https://cris.unibo.it/retrieve/handle/11585/701396/496403/FAIA313-0101.pdf - see the Annex, which looks to me like coding shoehorned into a metadata slot, rather than markup of text as such.

1 Like
#5

LRML can do more, just not alone.