Internet of Turtles 'No data to match against'

In which Wildbook did the issue occur?
Internet of Turtles
What operating system were you using? (eg. MacOS 10.15.3)
Windows 10
What web browser were you using? (eg. Chrome 79)
Chrome 94
What is your role on the site? (admin, researcher, etc)
Researcher, rest, machinelearning
What happened?

There should have been plenty of data/ encounters to match against. Also not sure why it would match only against 2 others in the same result. It is also still matching against this one encounter from Pensacola Beach Pier, which is not in Hawaii…Keeps popping up
What did you expect to happen?
Normal results to show up
What are some steps we could take to reproduce the issue?

If this is a bulk import report, send the spreadsheet to with the email subject line matching your bug report


This is still happening (also see above old message):

Some of these are re-sights where I know we have encounters for here on IOT.
Not sure what the problem is. I can ‘start another match’ but then I’m not sure if you can still see the issue once you click on the link

Hi, @noaahawaii !
I’ll likely get to this by the end of the week (Thursday). If it can wait until then, I’ll take a look. Otherwise, go ahead and “start another match”.

Thanks! I will hold off running matches until you have had a look. I will keep uploading encounters in the meantime

1 Like

@MarkF & @colin
Hi Mark, just following up on your message from 7 days ago. Any progress?


Hi, @noaahawaii !
I’m so sorry for the delay. I have to confess that I’m not quite sure what’s going on here.
Here’s some useful information that I’ve been able to glean so far (perhaps for @colin or @jason or someone else more familiar with the image analysis part of our product to inspect further):

HotSpotter seems to be working… but it’s match candidate set is very small (2).
Looking at Internet of Turtles, for instance, I can see in the logs that the only matchingSetFilter is locationId? Which should yield more than two candidates:
Searching the logs for the failed task id (b9169615-353d-4e52-8aff-1ece999b0c2e ) from that same results page yields:

catalina.out.1:+ starting ident task b9169615-353d-4e52-8aff-1ece999b0c2e

catalina.out.1:INFO: IAGateway.processQueueMessage() 'identify' from successful --> {"success":true,"taskId":"94bfbdcd-3cf4-4552-9905-52c59f54322c","tasks":[{"annotationIds":["5a9d88d0-d46f-4f49-940f-b98edf32ffa4"],"beginIdentify":{"sendAnnotationsAsNeeded":{"numAnnotsTotal":1,"numAnnotsToSend":0,"numAssetsToSend":0},"success":false,"_action":"error","error":{"emptyTargetAnnotations":true,"message":"rejected","error":"Empty target annotation list"}},"error":{"emptyTargetAnnotations":true,"message":"rejected","error":"Empty target annotation list"},"taskId":"b9169615-353d-4e52-8aff-1ece999b0c2e"},{"annotationIds":["ed1305cf-dd99-424d-b3f1-606998dade5d"],"beginIdentify":{"sendAnnotationsAsNeeded":{"numAnnotsTotal":2,"numAnnotsToSend":0,"numAssetsToSend":0},"sendIdentify":{"response":"41a2c482-6054-4cb4-8e9c-1b4019eea47c","status":{"cache":-1,"code":200,"success":true,"message":""}},"success":true},"taskId":"2c94e950-5a69-406b-9c46-011f7d560a0e"},{"annotationIds":["01dc1e07-c9f9-40ef-8a70-33b58759b531"],"beginIdentify":{"sendAnnotationsAsNeeded":{"numAnnotsTotal":2,"numAnnotsToSend":0,"numAssetsToSend":0},"sendIdentify":{"response":"6a6887c7-5fd8-440a-9d12-046a526b4e23","status":{"cache":-1,"code":200,"success":true,"message":""}},"success":true},"taskId":"5ddb5576-bbe4-4070-b002-82f673352a9e"}]}

localhost_access_log.2021-10-21.txt: - - [21/Oct/2021:17:28:47 +0000] "GET /iaLogs.jsp?taskId=b9169615-353d-4e52-8aff-1ece999b0c2e&projectId=null HTTP/1.0" 200 741

So the logs are confirming a “emptyTargetAnnotations” error.

I also noticed the encounter target for this match has a manual annotation: Internet of Turtles

However, this is not true of the encounter target for another failing match:
: Internet of Turtles.

I guess my questions to you @noaahawaii are: do you know what other search criteria that you used?

As in, which algorithms did you select in addition to hotspotter? Did you narrow the search by “my data”?

This is indeed pretty weird, especially the Pensacola Beach Pier candidate. It’s not weird that, if the match set finds itself with zero candidates, the search gets widened automatically to the whole dataset (i.e, "well, try to find something). But why would the search only be widened to include this one other candidate??

Perhaps of note is the fact that the Pensacola Beach Pier candidate (Internet of Turtles) does NOT have a locationId designated. But neither do 814 other encounters in the IoT database.

I’ll try to recruit some help for this.

I’m glad that there at least generally seems to be a workaround in the sense that “Start another match” seems to resolve the issue?

If these analyses are urgent for you in any way, I encourage you to just go ahead and kick off the matches again.

Otherwise, if you want to leave one or two in their current wonky state, that will be helpful for future troubleshooting.

I created a ticket for this that we’re tracking internally as WB-1845.


@MarkF @colin @jason
I did not use any other search criteria. These ran automatically after uploading when logged in. It does not give me an option at that point to select different criteria or algorithms.

Thanks for looking into this.

1 Like

Hi, @noaahawaii !
Thanks! We’re pursuing a lead with the O’ahu locationId being sanitized to Oahu somewhere in the pipeline.
Will report back when there’s more info!