Observations still marked as spam after resolved spam flag on account

Bug Reports
1

Platform Website

Browser, if a website issue: Chrome

I was having a hard time finding an observation of mine I was looking for in the Identify search, and stumbled across this issue. It may extend earlier than 2017, but that’s when my earliest observation uploads happened.

According to the My Observations list, between Jan 1 2017 and Feb 7 2018, I created 1633 observations:

32
32938×982 121 KB

URL: https://www.inaturalist.org/observations?created_d1=2017-01-01&created_d2=2018-02-07&place_id=any&subview=table&user_id=graysquirrel&verifiable=any

However, when I go to the Identify tab and do a search for the same time period (filters set to include any quality grade, any reviewed status, any place ID), I get exactly 2 results - and one of those only started to appear after I unmarked and re-marked its cultivated status (the manzanita).

31
31926×861 60.9 KB

URL: https://www.inaturalist.org/observations/identify?reviewed=any&quality_grade=casual%2Cneeds_id%2Cresearch&verifiable=any&user_id=graysquirrel&created_d1=2017-01-01&created_d2=2018-02-7&createdDateType=range&place_id=any

I’ve noticed that sometimes I’ll ID every observation of a species in my county, and then a few weeks later there’ll be more observations of that same species, supposedly posted years ago, that didn’t appear in my original search. I’m thinking that issue may be related to this one, and older observations just don’t always show up for some reason.

1 Like
2

I tried it using my observations, and got the same number (2,231) both ways. I was using Safari.

3

Interesting - I tried it with yours, and it worked fine. Can you try with my observations and see what results you get?

1 Like
4

I recreated your links (just to have it clear) and they’re working the same as you describe them. I did the same with mine but another period (and had to go for shorter one), all observations were shown.
I’d say regular id tab is wroking weird too, the order of “last updated” sometimes is ridiculous and newer observations go after many hundreds of old ones where nothing could be changed recently.

1 Like
5

I did a bit more fiddling around, and removed all the date restrictions on my observations. Seems a lot of the affected ones are in that range, but not all.

For my all-time verifiable observations, this shows 33,829: https://www.inaturalist.org/observations?place_id=any&subview=map&user_id=graysquirrel

Identify shows 30,638: https://www.inaturalist.org/observations/identify?reviewed=any&quality_grade=needs_id%2Cresearch&user_id=graysquirrel&place_id=any

Which incidentally is almost the same discrepancy as the one between number of observations listed in Explore leaderboards (30,628): https://www.inaturalist.org/observations?place_id=any&subview=map&view=observers
vs the number shown on my profile (33,847): https://www.inaturalist.org/people/graysquirrel

Either way, that’s over 3,000 observations that nobody can identify for me because they don’t show up in searches… I really hope I don’t have to manually go through them all and mark/unmark as “not wild” on every one just to make it appear.

6

I don’t have any comment for why this happened, but the missing observations have spam=true, so they’re automatically filtered out of Identify.

Compare the results of
https://api.inaturalist.org/v1/observations?quality_grade=casual%2Cneeds_id%2Cresearch&verifiable=any&user_id=graysquirrel&created_d1=2017-01-01&created_d2=2018-02-7&spam=true

and

https://api.inaturalist.org/v1/observations?quality_grade=casual%2Cneeds_id%2Cresearch&verifiable=any&user_id=graysquirrel&created_d1=2017-01-01&created_d2=2018-02-7&spam=false

7

Ooh interesting - a few days ago someone apparently accidentally flagged my user account as spam and then restored it. I have a suspicion that may have caused this somehow.

4 Likes
8

I changed the title to make it a bit clearer what the issue is

2 Likes
9

embarrassingly enough that this happened in the first place (sorry again!), but I feel really bad there seem to be further problems.

The flag is this one: https://www.inaturalist.org/flags/557497 - that might help solve the problem.

Maybe the system started flagging observations as spam until I resolved the problem, but then somehow did not change back?

1 Like
11

Not to worry, it was bound to happen to someone eventually, and at least this way it gets noticed and hopefully resolved :) I think it’s a bug - as I understand it, flagging a profile as spam shouldn’t affect observations from that profile unless the observations themselves are also flagged - @tiwane can you confirm if my understanding is correct here?

And that’s assuming that IS even the cause, it could be just coincidental and caused by something else entirely.

1 Like
12

(just tested this on our test server)

If an account is flagged as a spammer, their observations can be found in Explore and, if one is logged in, the individual observations can be viewed. If one isn’t logged in, one can’t view a spammer’s individual observations.

13

Can they also be found in Identify?
Presently I have 30,804 verifiable observations available in Identify and 33,985 in Explore

EDIT: If I locate one of the missing observations and change the DQA in some way, like marking and then unmarking as “cultivated” it starts to reappear in identify searches. Is there some way that can be accomplished without me having to go through 30k+ observations to find them and fix them?

2 Likes
14

You can search for a spammer’s observations in Identify but none of them will appear.

EDIT: talking it over, it’s likely that allowing observations from an account flagged as a spammer to show up in Explore is probably an oversight. We’ll probably end up restricting that.

I’m pretty sure we can get those reindexed, I’ll ask our devs to be sure, though.

2 Likes
15

OK, they’ve been reindexed. How’s it looking?

2 Likes
16

They’re all back where they should be! Thank you so much!

1 Like
18

Filed a bug report for this issue: https://github.com/inaturalist/inaturalist/issues/3348

2 Likes