[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/vt/ - Virtual Youtubers

Search:


View post   

>> No.31031319 [View]
File: 122 KB, 680x799, 1636795775286.png [View same] [iqdb] [saucenao] [google]
31031319

It's actually totally possible to use machine learning and NLP to identify posts in the thread automatically.
But you need a decent amount of data (messages) from each individual with 100% certainty. For vtubers you have far enough with twitter, clear selfposts and discord chats. But for viewers you're limited to their posts in discords or twitch chat, and the way of writing on the board can be different so there is a margin of error.

Nevertheless it would be possible to apply it to create a writing style profile for each person and then for each anonymous post have the list of the highest % of probability authors.

I'm not sure if using the identified posts as data itself would be good (to make it closer to the specifics of that person's way of writing on the board vs on twitter) or would be too risky in case of errors as it would reinforce its mistake more and more.

Navigation
View posts[+24][+48][+96]