0.25
0.5
0.75
1.25
1.5
1.75
2
Adding value to NLP: a little semantics goes a long way.
Published on 2019-07-1964 Views
Natural language processing technology is now ubiquitous, even if there are still many challenges to be faced in its development. From sentiment analysis to machine translation to chatbots; from medic
Related categories
Presentation
Adding value to NLP: a little semantics goes a long way 00:00
15 years ago in NLP, we were00:13
NLP at ESWS 200401:34
Mapping semantic relations02:17
Merging descriptions of plants in Multiflora02:49
Semantic annotation 15 years ago at Ontotext03:23
And at the OU03:41
Many of the underlying technologies are still relevant03:52
How to impress a “freedom of the media” researcher06:11
Feral databases06:49
That’s great, we have lots of NLP tools to do that!07:43
Woe betide anyone who opens the sacred spreadsheet!08:04
Harry Potter and the Spreadsheet of Doom08:38
Once the latest in calendar technology09:13
Case studies: how can NLP and semantics help?09:47
Global monitoring of violations against journalists09:56
Events are complex12:14
But we know how to understand complex events13:08
Information categories in current monitoring efforts13:53
ICCS Crime Classification Scheme14:56
HURIDOCS classification scheme15:12
Putting it all together15:43
Reconciling information from different sources17:04
Reconciling our information18:24
Automatic categorization of free text 18:44
Social Media and Disaster Relief19:00
Aid workers in Nepal discussing strategy 19:57
Tools to help disaster victims get aid quickly: pinpointing geographic locations20:44
CREES Google Sheets Add-on22:13
Semantic Technologies in Scientometrics23:04
What kind of questions do we want to answer?24:59
Which countries published most about waste management and recycling in 2014?25:27
All publications25:53
How is European knowledge distributed across regions?26:52
Technological vs scientific knowledge production in genomics27:40
Specialisation Indexes in Biotechnology around Europe28:25
The Semantic Approach28:50
Ontologies connect information29:10
From ontology to data29:39
Creating and populating the ontology30:28
SGC Topics and SubTopics30:44
Linking information from external sources31:01
Linking information from external sources - 231:08
Ontology population31:34
Annotating Data with Ontologies32:05
Annotating Data with Ontologies - 232:49
Example of patent annotation33:13
Ongoing Challenges33:38
GATE MIMIR and Prospector34:35
GATE MIMIR and Prospector - 235:19
Prospecting Biomedical Literature36:37
Choosing A Specific Instance37:04
What diseases are in these documents?37:11
What pathogens?37:29
Disease vs Disease Co-ocurrences37:34
Diseases vs Pathogens38:06
Social media and politics38:27
Hate speech against politicians Jan-Feb 201938:58
Who gets the abuse?39:47
Top abusive words41:07
Words vs people41:40
Topics of abuse per party42:17
Local and national news coverage of Brexit43:24
So where are we at?46:06
Acknowledgements48:10