communitynotes icon indicating copy to clipboard operation
communitynotes copied to clipboard

Abnormal tweetId in Note request data

Open avalanchesiqi opened this issue 4 months ago • 0 comments

The newly released note request data is said to have four columns "userId tweetId createdAtMillis sourceLink". I notice that a non-trivial amount of tweetId seem to be abnormal.

Example:

cat batSignals-00000.tsv| grep $'\t91540'
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	9154089611629224624	1724916211485
BE177514EB62FA0F486A310A3DD07698C591ACE5077E70E5B5E9AFC105DD2DCE	9154092414728527376	1726533818655

Usually, tweetId starts with "1" but the above two tweetIds start with "9". I also searched for all requests proposed by the first userId.

cat batSignals-00000.tsv| grep C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	977040036244765513	1724916298043
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	1828757772290400291	1724935602779
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	2684673839852374649	1724319335350
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	2951003568237083342	1725515064463
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	3324343820942848469	1724258023892
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	3561937426136355544	1724743384355
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	4137635619809309092	1724258370090
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	4952653969869310800	1724936298653
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	5651508576841648976	1725515613590
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	5724641268837992313	1724326146100
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	6905583205459998959	1724916234226
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	6985250573715648597	1724257857799
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	7195974708253445253	1724126266603
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	7538073314496696227	1724937322549
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	7820944466784416933	1725108506437
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	8558507768692594852	1725108335051
C23DBB56A37D9566AE659462EA431791A55D9403F8E0B54E5E5899E236B9526A	9154089611629224624	1724916211485

It looks like all of this person's requests are on some sort of abnormal tweetIds. All of these tweets are not available on X. Is there a bug in exporting tweetIds for the note request data?

Updated: The requests with abnormal tweetIds are very small fraction---about 0.09%.

avalanchesiqi avatar Sep 10 '25 13:09 avalanchesiqi