-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
how does offline name-finding (non-verification) work? #122
Comments
Name finding uses dictionaries and heuristic rules as the first pass, Bayes as a second pass. So if a name is simple and has a misspelling, usually it will be ignored. However, if Bayes algorithms collect enough "points" they will register a name candidate. The rules are relaxed, and verification is an important step to weed out name-like combinations. |
Internet connection is not needed if |
I am wondering how the
"verification": false
option works.I would expect that gnfinder checks words which look like known Latin/latinized names or epithets, and "guesses" they are part of a scientific name.
So I tried to change one letter in names (i.e. from Quercus toza to Quercus tozza in example #121), to check if they were returned as fuzzy matches (with
verification=true
and then withverification=false
)Although Quercus toza is found to be a scientific name in both cases, Quercus tozza is not (and gnfinder returns just "Quercus" genus match).
I can understand a genus-only match when
verification=true
if for some reason the fuzzy algorithm was not able to match toza to tozza.But with
verification=false
, I was expecting gnfinder to find anything that "looks like" a scientific name, even if it was never published, just by looking at its separate words as being part of other names (i.e., "Homo sylvestris")So ... what should I expect with
verify=false
? How does gnfinder make decissions in that case?If no internet connection is used, why does gnfinder say that Quercus toza is a valid name, but Quercus tozza is not?
It looks like as if names are being "verified" against name sources anyway (although no verification is output in json result).
The text was updated successfully, but these errors were encountered: