0.25
0.5
0.75
1.25
1.5
1.75
2
Research at Twitter
Published on Nov 04, 201310007 Views
Twitter is a company dealing with broadcasting messages across the globe in real-time. In this talk we will make an overview on how Twiiter operates in terms of technical infrastructure, tools and r
Related categories
Chapter list
Research at Twitter00:00
Los Angeles Earthquake02:40
#ocws, #occupywallstreet, ....04:08
The Scale of Twitter05:10
Internal vs. External R&D06:39
Overview08:06
Problems (1)08:43
Problems (2)10:07
Problems (3)11:15
Personalization11:54
Key Research12:34
Graph mining13:02
Follow graph (1)13:31
Follow graph (2)15:01
wtf15:41
"Circle of Trust"17:41
SALSA (1)18:53
SALSA (2)20:25
Engagement21:15
Heterogeneous22:23
User/content categorization23:05
Interest modeling23:16
Challenges24:52
User topic attribution25:32
Tweet topic attribution26:52
Integration28:50
Multi-label classification at scale (1)30:07
Multi-label classification at scale (2)31:35
Chatter detection (1)32:35
Chatter detection (2)33:47
Topical vs chatter34:13
Chatter detection (1)35:15
Chatter detection (2)36:02
Internattional support37:18
Bootstrapping38:31
Spammy interactions39:39
Spam41:18
Adversarial games42:12
Tweet level42:20
Tweet-history level43:08
Account sales43:56
Pricing44:55
Legitimacy signals45:07
Big data vs Fast data47:18
Hashtag topic tracking48:10
Human computation (1)50:43
Human computation (2)52:43
Plumbing53:20
Scalable analytics (1)54:36
Scalable analytics (2)56:36
Scalable analytics (3)57:40
Scalable machine learning58:30
Models59:10
Model training59:54
Supervised classification (1)01:01:34
Supervised classification (2)01:02:23
Ensembles (1)01:04:04
Ensembles (2)01:05:11
Example (1)01:06:07
Example (2)01:07:01
Stream-based machine learning01:08:49
Big boxes01:11:05
Thank you01:13:17