problem with citations

[legaltextai]

take e.g. this case from the latest 'citations' file:
cluster_id 13914752 : 488 U.S. 984
the citation shows Levitt v Univ of Texas according to Google Scholar https://scholar.google.com/scholar?hl=en&as_sdt=800006&q=488%09U.S.%09984&btnG=
It should be Pollard v Owens according to CL web search https://www.courtlistener.com/opinion/9083109/pollard-v-owens-illinois-inc/
and it is Armentero v. Makel if you search for this cluster_id in the bulk files (opinions+cluster+docket)
i have found several of these mismatches
is 'cluster_id' the same in both opinion_cluster and citations files?
what is the most reliable source for citations? and if it is not by cluster_id , any recommendation how else to merge citations with opinions?
thank you!

[legaltextai]

Not sure I understand. When I join by cluster_id I run into the issue I described. If that’s not the way to merge the files, how do I connect the name of the case to its proper citation?

…

On Thu, May 2, 2024 at 2:43 PM Mike Lissner ***@***.***> wrote: I think the answer you're looking for is that the bulk citations file contains links between opinions. Does that fix these problems? — Reply to this email directly, view it on GitHub <#4012 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BCNHY22E33KNOMYNHPEZY43ZAKCOVAVCNFSM6AAAAABHEG24W2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJRGI2TSNZWGA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[mlissner]

My first response misunderstood your question, so I hid it.

---

488 U.S. 984 can be several cases:

https://cite.case.law/us/488/984/

This is because each of the cases is just one or two lines long and the publisher put several on the same page of the book.

When I go to the cluster ID you provided:

https://www.courtlistener.com/opinion/13914752/x/

It throws a 404, so that's weird.

is 'cluster_id' the same in both opinion_cluster and citations files?

Yes.

So ultimately, if you want to merge a citation with a cluster, you'll want to join the cluster_id in the citation object with the id field of the cluster object.

[legaltextai]

yeah, I did not realize several cases may be on the same page in the reporter. I excluded all opinions < 200 words, ran a random check on citations, looks fine now.