New iPhone app: CV suggestions not considering location

[timo27]

Platform: iNat Next (+Website)

App version number, if a mobile app issue: 1.0.5 (168)

Browser, if a website issue:

URLs (aka web addresses) of any relevant observations or pages: https://www.inaturalist.org/observations/296635031

Screenshots of what you are seeing:

Description of problem:

Step 1: open the observation and attempt to add an id. Look at what the CV suggests.

Step 2: open the same observation but on the iNat Next app and attempt to add an id. On the app, it suggests something very different.

It seems the CV fails to access the location of the observation on the new iPhone app. Coelophora saucia is an East-Asian species that doesn’t occur anywhere near Switzerland. The CV for the website works though.

(PS The observation was originally created on iNat Next and it originally also suggested Coelophora saucia, but i chose Tribe Coccinellini just to be safe)

(PPS restarting the app doesn’t fix the issue, nor does restarting the phone, nor does logging out and in, nor does uninstall/reinstalling)

[tiwane]

Hmm, I can replicate your results in the app and on the website. Not sure what’s going on, but we can take a look. Are you seeing other cases of this?

[timo27]

So i found out that the CV on the app does take the location into account, albeit in a lesser-weighted manner compared to the website.

Website (weak/strong suggestion rating derived from Chrome extension):

• NO location:
  We’re pretty sure this is in the genus: Genus Coelophora (strong)
  Top suggestion: Coelophora saucia (strong)

• WITH location:
  We’re pretty sure this is in the tribe: Black-spotted Lady Beetles (strong)
  Top suggestion: Harmonia axyridis (weak) (correct)

App (iNat Next):

• NO location:
  Top ID suggestion: Coelophora saucia (87.2%)
  Other suggestion: Oenopia formosana (4.2%)
  Other suggestion (2nd): Harmonia axyridis (0.8%) (correct)

• WITH location:
  Top ID suggestion: Black-spotted Lady Beetles (83.8%)
  Other suggestion: Coelophora saucia (53.5%)
  Other suggestion (2nd): Harmonia axyridis (17.4%) (correct)

All using the same image and same location.

It’s worth noting the biggest difference: provided the location, the website doesn’t at all suggest Coelophora saucia, whereas the app suggests it 53.5%

[eyekosaeder]

I too have noticed the app suggesting inaccurate species IDs with species that do not occur here. Not this particular case, but I’ve seen a bunch of North American species that have no observations nearby.

FWIW, H. axyridis f. conspicua relatively frequently get IDed as C. saucia (even before the new app)