0.25
0.5
0.75
1.25
1.5
1.75
2
Mining Topics in Documents: Standing on the Shoulders of Big Data
Published on Oct 08, 20143704 Views
Topic modeling has been widely used to mine topics from documents. However, a key weakness of topic modeling is that it needs a large amount of data (e.g., thousands of documents) to provide reliable
Related categories
Chapter list
Mining Topics in Documents Standing on the Shoulders of Big Data00:00
topic models require a large amountof docs00:08
Example Task Application00:18
Can we improve modelingusing Big Data?00:29
Human Learning01:00
Motivation01:07
Proposed Model Flow01:39
What’s the knowledge representation?02:13
How does a baby gain knowledge?02:29
Knowledge Representation02:44
Knowledge Extraction02:49
Frequent ItemsetMining (FIM)03:05
Extracting Cannot-Links03:23
Related Work about Cannot-Links03:55
However, both of them assume...04:12
Knowledge Verification04:49
Must-Link Graph05:08
PointwiseMutual Information05:21
Cannot-Links Verification05:46
Proposed Gibbs Sampler06:25
Example06:40
M-GPU - 107:04
M-GPU - 207:21
M-GPU - 307:42
M-GPU - 407:59
M-GPU - 508:06
M-GPU - 608:11
M-GPU - 708:20
M-GPU - 808:31
Evaluation08:52
Model Comparison09:32
Topic Coherence09:53
Topic Coherence Results10:14
Human Evaluation Results10:53
Electronics vs. Non-Electronics11:50
Conclusions12:50
Future Work13:20
Q&A14:02