Pascal Challenge on Evaluating Machine Learning for Information Extraction: Goals, Results and Conclusions

author: Neil Ireson, Web Intelligence Technologies group, University of Sheffield, Department of Computer Science, Department of Molecular Biology and Biotechnology, University of Sheffield

Categories

Top: Computer Science: Information Extraction

You might be experiencing some problems with Your Video player.

*Slides*
0:00	PASCAL CHALLENGE ON INFORMATION EXTRACTION & MACHINE LEARNING
0:04	Organisers
0:10	Outline
1:01	Goal : Provide a testbed for comparative evaluation of ML-based IE
1:53	Data (Workshop CFP)
3:42	Data (Workshop CFP)
4:27	Data (Workshop CFP)
4:46	Data (Workshop CFP)
5:45	Preprocessing
6:00	Annotation Exercise
7:00	Annotation Document - sample
9:47	Annotation Slots
11:01	Evaluation Tasks
12:11	Evaluation
12:33	Participants
13:52	Task1
13:55	Task1: Test Corpus
14:09	Task1: Test Corpus
15:39	Task1: 4-Fold Cross-validation
16:23	Task1: 4-Fold & Test Corpus
17:41	Task1: Slot FMeasure
19:51	Best Slot FMeasures Task1: Test Corpus
20:46	Task 2a
20:53	Task2a: Learning Curve FMeasure
22:34	Task2a: Learning Curve Precision
23:00	Task2a: Learning Curve Recall
23:13	Task 2b
23:19	Active Learning (1)
23:49	Active Learning (1)
24:09	Active Learning (2)
24:41	Active Learning (2)
24:55	Active Learning (3)
25:02	Active Learning (3)
25:48	Task2b: Active Learning
26:43	Task2b: Active Learning Increased FMeasure over random selection
28:27	Task 3
28:43	Conclusions (Task1)
29:49	Conclusions (Task1: Test Corpus)
30:19	Conclusions (Task1: 4-fold Corpus)
31:01	Conclusions (Task1)
31:40	Conclusion (Task2 & Task3)
32:25	Discussion

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.