Useless Email Things: Difference between revisions

Revision as of 11:15, 29 May 2022

Weeding out useless transmissions in a sample base from 280,000 emails via three accounts back to roughly 2006.

One downside of my approach is that I'm not capturing a high level count of things before removing, which skews some of the output numbers as we go: if an email is found earlier by a search term in removed, it can't be counted in subsequent searches.
I had to stop processing Facebook things for a bit. I could just delete them all, however there are a bunch of conversations and birthdates and... nostalgia... ugh

From\To	Subject	Count	Notes
[email protected]		12265	Useless Twitter Emails
[email protected]	check out the notifications you have on Twitter	4719
[email protected]	is now following you on Twitter	200	boy, that's sad
[email protected]		832
	%	9239	there were ten legit emails in there, seven of which were really not interesting
	$	11964	I kept 46 due to nostalgia. 200 or so were from marketplace or craiglist interactions (man are those painful). Newegg, Priceline, and the like... 1000's of them.
	"iron force"	2100	some game I played
promo@		3915	I had meant to go @e.newegg and @newegg.com, but promo hit all new eggs stuff. Spiffy!
@facebookmail.com	added you	230
@facebookmail.com	wrote on your wall	400
@facebookmail.com	invited	689
@facebookmail.com	sent you a message	1224	"sent you a message." marks the start of the OPEN IN MESSENGER crapfest. "*... on Facebook" contains the actual message.
@facebookmail.com	replied to a thread	100
@facebookmail.com	suggested	400
@facebookmail.com	birthday	4835	that's going to require a lot of cake...
@facebookmail.com	commented	20874	between Aug 2020 and June 2021, the switch to Open In Messenger. Older ones have the comment inline.

@@ Line 1: / Line 1: @@
-Weeding out useless transmissions in a sample base from 280,000 emails via three accounts back to roughly 2006.  One downside of my approach is that I'm not capturing a high level count of things before removing, which skews some of the output numbers as we go: if an email is found earlier by a search term in removed, it can't be counted in subsequent searches.
+Weeding out useless transmissions in a sample base from 280,000 emails via three accounts back to roughly 2006.
+* One downside of my approach is that I'm not capturing a high level count of things before removing, which skews some of the output numbers as we go: if an email is found earlier by a search term in removed, it can't be counted in subsequent searches.
+* I had to stop processing Facebook things for a bit. I could just delete them all, however there are a bunch of conversations and birthdates and... nostalgia... ugh