Discovering Temporal Bisociations for Linking Concepts over Time

Published on 2011-11-302001 Views

Corrado Loglisci

Bisociations represent interesting relationships between seemingly unconnected concepts from two or more contexts. Most of the existing approaches that permit the discovery of bisociations from data r

Sessions

Presentation

Discovering Temporal Bisociations for Linking Concepts over Time00:00

Bisociations - 100:17

Bisociations - 200:48

Computational Challenge01:19

The task: Link Discovery02:15

The task: Issues02:53

How to represent the contexts? - 103:08

How to represent the contexts? - 203:27

How to represent the contexts? - 304:45

How to represent the contexts? - 406:16

How to represent the contexts? - 507:14

How to perform link discovery? - 107:40

How to perform link discovery? - 209:06

How to perform link discovery? - 309:26

How to perform the link discovery process - 109:33

How to perform link discovery? - 409:56

How to perform link discovery? - 510:45

How to perform link discovery? - 611:33

How to perform link discovery? - 711:53

Application12:32

Application: pre-processing13:34

Results: running times14:05

Results - 115:14

Results - 215:53

Results - 316:35

Results - 417:10

Conclusions & Future Work17:10

Questions?18:01

Results: simulating *Swanson’s discoveries18:24

Results: X= Migraine, Y=Magnesium Deficiency - 118:39