en
0.25
0.5
0.75
1.25
1.5
1.75
2
SigniTrend: Scalable Detection of Emerging Topics in Textual Streams by Hashed Significance Thresholds
Published on Oct 07, 20143093 Views
Social media such as Twitter or weblogs are a popular source for live textual data. Much of this popularity is due to the fast rate at which this data arrives, and there are a number of global events
Related categories
Chapter list
SigniTrend: Scalable Detection of Emerging Topics in Textual Streams by Hashed Significance Thresholds00:00
Trend detection on streams should be early and accurate - 100:14
Trend detection on streams should be early and accurate - 200:51
Trend detection on streams should be early and accurate - 300:56
Trend detection on streams should be early and accurate - 401:18
Trend detection on streams should be early and accurate - 501:42
Problem description02:05
SigniTrend on textual streams03:13
Preprocessing03:52
Trend detection cycle - 104:50
Trend detection cycle - 205:02
Trend detection cycle - 305:11
Trend detection cycle - 405:18
Update statistics - 105:32
Update statistics - 205:59
Significance and frequency for term “Facebook”06:46
How to track statistics of all pairs efficiently?07:05
Hashing scheme for efficient tracking - 107:57
Hashing scheme for efficient tracking - 208:12
Hashing scheme for efficient tracking - 308:15
Hashing scheme for efficient tracking - 408:29
Hashing scheme for efficient tracking - 508:44
Hashing scheme for efficient tracking - 609:01
Hashing scheme for efficient tracking - 709:14
Hashing scheme for efficient tracking - 809:29
Artificial trends evaluation10:28
Refinement & clustering11:47
Conclusion12:43
Thank you!13:31