****s, need your help for a research project

[Top_Cat]

Developing a ensemble classifier for a machine learning project and need help from people. Basically I'll be doing a statistical labelling exercise using Conditional Random Field (CRF) models trained using the Enron email corpus to be used in a further bit of unsupervised AI/learning and need crowd-sourced suggestions for stuff to tag bit of emails.

I need examples of overuse of jargon, TLA's (Three-Letter Acronyms), clearly non-work stuff, phrases which people tend to say under stress, etc. the more ridiculous the better. In return, I offer free hugs and credits if/when some loser AI/ML/computational journal publishes it.

So let's hear it Facebook, what are some of the dumbest things you've seen in emails?

 

[sledger]

WAC is one obvious one that springs to mind.

 

[HeathDavisSpeed]

NFI mate. I can't help at all.

 

[Ausage]

WAC is one obvious one that springs to mind.

It seems like the thread was created solely for this purpose.

TMI is my actual response to the thread.

 

[Spikey]

fyi lol

 

[ajdude]

So let's hear it Facebook,

hey y'all don't say that

 

[Dan]

The ones used 'under stress' the most would be along the lines of WTF or OMG, surely?

Also have a couple of friends who use ONG after a misspelling stuck, or simply mash the keyboard in frustration.

 

[ajdude]

fmd

 

[ajdude]

isr

 

[Ausage]

Ikr.

Lol is one that gets chronically overused lol.

 

[G.I.Joe]

FFS.

Never seen it used outside CW, FTR.

 

[Agent Nationaux]

Uve - as in Uve is a ****.

 

[sledger]

GFC

z

 

[Riggins]

while we're at it could someone do my research lit review for me pls.

 

[sledger]

What's it on?

 

[Riggins]

graphene binding peptides.

 

[Ausage]

Get ready for a BFA!

 

[Jono]

Developing a ensemble classifier for a machine learning project and need help from people. Basically I'll be doing a statistical labelling exercise using Conditional Random Field (CRF) models trained using the Enron email corpus to be used in a further bit of unsupervised AI/learning and need crowd-sourced suggestions for stuff to tag bit of emails.

This better be a piss-take.

 

[Top_Cat]

'tis but it'll seriously be put up for peer-review because the ML methods are legit.

 

[Red Hill]

I have NFI what this is all about. FMD.