What Makes Writing Great? First Experiments on Article Quality Prediction in the Science Journalism Domain  thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

What Makes Writing Great? First Experiments on Article Quality Prediction in the Science Journalism Domain

Published on Oct 02, 20133855 Views

Great writing is rare and highly admired. Readers seek out articles that are beautifully written, informative and entertaining. Yet information-access technologies lack capabilities for predicting art

Related categories

Chapter list

What makes Writing Great? First experiments on Article Quality in the Science Journalism Domain00:00
A new task00:25
This work01:21
Science journalism genre01:58
Corpus of article quality03:00
Category 1 : VERY GOOD articles03:07
Expanding the VERY GOOD set04:36
Category 2: TYPICAL writing in the NYT05:26
A topic-paired corpus06:08
Two quality prediction tasks07:04
Properties of the dataset07:27
How to predict interesting writing?08:10
Approach to feature development08:17
1. Unusual words and phrases09:32
2. Visual nature10:16
Visual nature10:55
Human interest and text structure11:41
Sentiment and Research12:15
Strengths in predicting article quality12:56
How the features vary13:05
Accuracies on the two tasks13:48
Combining reader interest with other aspects14:22
Comparing with a bag of words system15:38
A word based classifier16:24
Accuracy of the Bag of Words system16:33
Conclusions17:29
Thank you17:57