29
Exposing Digital Content as Linked Data, and Linking them using StoryBlink Ben De Meester Tom De Nies, Laurens De Vocht, Ruben Verborgh, Erik Mannens, and Rik Van de Walle University Ghent – iMinds – Multimedia Lab [email protected] | @Ben__DM NLPDBpedia2015@ISWC | October 11 th 2015 | Bethlehem, PA

NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Embed Size (px)

Citation preview

Page 1: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Exposing Digital Content as Linked Data,and Linking them using StoryBlink

Ben De MeesterTom De Nies, Laurens De Vocht,

Ruben Verborgh, Erik Mannens, and Rik Van de Walle

University Ghent – iMinds – Multimedia [email protected] | @Ben__DM

NLPDBpedia2015@ISWC | October 11th 2015 | Bethlehem, PA

Page 2: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

We live in a fast worldwith a lot of content to sift through

http://blog.qmee.com/qmee-online-in-60-seconds/

Page 3: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Book ≠ Fast

Page 4: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Finding a good book in short time?

Recommendations!

Page 5: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Recommendations?

Social recommendationsLong tail

Metadata recommendationsManual?

Page 6: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

What do we want?

Automatic content-based metadata

to fuel future recommendation-engines

Page 7: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Content-based metadata

Get the tags…DBPedia Spotlight

... use them to represent books’ content …EPUB CFI, NIF, ITS, …

… and link to other books … in a good way.TPF, EiCE

Storyblink!

Page 8: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Get the tags

Find out what a book is about…

Semantic tags!

Using NER/NED!

Extract all semantic concepts from the book

Page 9: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

AGDISTIS

Page 10: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

AGDISTIS

Open source

Local

NER/NED/NEL

Page 11: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

From a book to a semantic book

… …

Split HTML into chunks

HTMLto text

Local Spotlight

Page 12: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Represent a book by tags@prefix schema: <http://schema.org/> .@prefix nif: <http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core#> .@prefix itsrdf: <http://www.w3.org/2005/11/its/rdf#> .@prefix dbr: <http://dbpedia.org/resource/> .@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> . @prefix pg84: <http://www.gutenberg.org/ebooks/84.epub#> .

pg84:book a schema:Book .

pg84:epubcfi(/6/12!/4/2/4) itsrdf:taIdentRef dbr:Chamois ; nif:sourceUrl pg84:book .pg84:epubcfi(/6/2!/4/46[chap01]/16/42) itsrdf:taIdentRef dbr:Chamois ; nif:sourceUrl pg84:book . pg84:epubcfi(/6/12!/4/2/6) itsrdf:taIdentRef dbr:Desert ; nif:sourceUrl pg84:book .

...

Page 13: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Represent a book by tags@prefix schema: <http://schema.org/> .@prefix nif: <http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core#> .@prefix itsrdf: <http://www.w3.org/2005/11/its/rdf#> .@prefix dbr: <http://dbpedia.org/resource/> .@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> . @prefix pg84: <http://www.gutenberg.org/ebooks/84.epub#> .

pg84:book a schema:Book .

pg84:epubcfi(/6/12!/4/2/4) itsrdf:taIdentRef dbr:Chamois ; nif:sourceUrl pg84:book .pg84:epubcfi(/6/2!/4/46[chap01]/16/42) itsrdf:taIdentRef dbr:Chamois ; nif:sourceUrl pg84:book . pg84:epubcfi(/6/12!/4/2/6) itsrdf:taIdentRef dbr:Desert ; nif:sourceUrl pg84:book .

...

Page 14: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Link to other books

Open Source

Linked data path finding

Multiple paths

Page 15: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink
Page 16: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink
Page 17: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink
Page 18: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink
Page 19: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink
Page 20: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Keeping all concepts…

Not all mentioned concepts are useful.

The path finding becomes really slow.

Page 21: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Keeping all concepts…

Not all mentioned concepts are useful.

The path finding becomes really slow.

What happens if we keep the top X%?

Page 22: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

0 10 20 30 40 50 60 70 80 90 1000

2

4

6

8

10

12

14

0

10000

20000

30000

40000

50000

60000

Amount of considered concepts (%)

#paths Time (s)

Top 50% of found concepts gives similar paths,but a lot faster

Page 23: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

0 10 20 30 40 50 60 70 80 90 1000

2

4

6

8

10

12

14

0

10000

20000

30000

40000

50000

60000

Amount of considered concepts (%)

#paths Time (s)

Top 50% of found concepts gives similar paths,but a lot faster

Time-out

Page 24: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Optimized Results@prefix schema: <http://schema.org/> .@prefix nif: <http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core#> .@prefix itsrdf: <http://www.w3.org/2005/11/its/rdf#> .@prefix dbr: <http://dbpedia.org/resource/> .@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .@prefix pg84: <http://www.gutenberg.org/ebooks/84.epub#> . pg84:book a schema:Book .

pg84:book itsrdf:taIdentRef dbr:Chamois, dbr:Desert, ...

http://uvdt.test.iminds.be/storyblinkdata/books

Page 25: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Storyblink

Exploring the links between classic works

Choose two books, and…

Page 26: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Storyblink

Page 27: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Next steps

Scale

Indirect pathse.g. book about WWI and book about WWII

Relevancy measuresKnowledge base influenceFiltering influence

Page 28: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Storyblinkgives a semantic representationof important semantic concepts

inside books, and uses those to connect books together content-wise

http://uvdt.test.iminds.be/storyblink

Demo 48

Page 29: NLP&DBpedia2015 - Exposing Digital Content as Linked Data, and Linking them using StoryBlink

Our project

The Publisher of the Future

Our pilot project partners: