[11:00:44] o/ [11:00:50] Hack day starts early for me. [11:00:53] jetlag! [13:43:14] halfak: about usernames; [13:43:14] I wanna solve (at least try) this problem myself. but I stuck at feature extraction. What can be the meaningful features of 1 word? [15:05:33] I think we should (1) explore badword detection within word and implement word-breaking strategies like splitting the word at underscores, capitalization, punctuation, etc. [15:05:53] mavrikant, ^ [15:05:57] Sorry I missed your ping. [15:11:30] mavrikant, we should be able to detect "Somewordthingsomeotherwords" but we'll have false positives. That should be OK. The machine can learn ways to operate around the falsepositives. [15:12:56] E.g. "PublishIt" contains "shIt", but it also, two clear dictionary words account for all of the characters in the string. I bet a learning model will note that duly. [16:23:13] o/ [16:23:29] I'm around for hacking today but I'm mostly focusing on paper review for CSCW 2018 [16:23:31] ugh. [16:23:36] So many reviews all at once. [16:26:57] halfak: good luck! [16:27:00] :) [16:27:05] Godspeed [16:28:34] thanks [16:28:46] Painful, yet important work. [16:29:23] "painful, yet important work" this does apply to me right now as well ;) [16:29:42] does apply to what I'm working on right now* [16:37:09] Hmm... I guess? The world would be easier if everyone knew how to do everything and I could focus on myself. Might be less fun though. [16:37:20] Cutting firewood is also painful yet important work. [16:37:36] But I'm not sure I want the firewood to cut itself. [17:15:52] OK off to do yardwork. [17:15:53] o/