API case_name incomplete? Party names

[nicoleprause]

I have a reasonable name-based gender classifier I would like to run on party names. However, the returns case_name and case_name_full appear to only have one party name that is not identified (as plaintiff, attorney name, etc.).

Am I missing party name data somewhere in the returned structure? Or does someone know a dataset I could join (e.g., based on docket number) that contains party name information?

[mlissner]

Can you please provide an example?

[nicoleprause]

When I use R to push my query to the API, I get three variables that appear to contain party name information:
case_name_full,case_name,case_name_short

Sample data from each variable:
case_name_full is empty
case_name: "Hilbert v. Uber Technologies, Inc.", "In re J.M. CA5", "In re R.K. CA6","Lopez v. County of Los Angeles CA2/1","Marriage of Arora CA6","People v. Bowser CA5"
case_name_short: "","Andres-Ordones","Diaz-Moreno","Fernandez","Funes-Pavon","Gonzalez-Irigolla"

My gender classifier relies on first names. Case_name is the closest of these three variables, but none contain first names and some have only one name of unclear role.

The query parameters are below. Incidentally, it appears the search is not working (e.g., https://casetext.com/case/united-states-v-vasquez-540, "United States v. Salsenio-Vasquez" contains none of the query words). I'll look at that next, but that depends if I can get first name data from this repository first.

query_params <- list(
  q = "defamation OR libel OR slander", 
  date_filed_min = "2019-01-01",  
  date_filed_max = "2025-01-01",  
  status = paste(
    c(
      "stat_Published",      
      "stat_Unpublished",   
      "stat_Errata",         
      "stat_Separate",     
      "stat_In-chambers",   
      "stat_Relating-to",   
      "stat_Unknown"       
    ),
    collapse = ","
  )
)

[mlissner]

Not sure what you mean with the casetext link?

Other than that, yeah, getting first names might be hard. They're normally omitted from the case name itself, but I guess they might sometimes appear in the case_name_full field. Alas, that's not usually populated though.

That leaves parsing the text of the decision, which I think you could probably do using AI, but it'd be tricky, I fear.