Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Exhibition recommendation 
using 
British Museum data 
and 
Event Registry
Introduction 
• Group 1: Ninja Turtles 
• Kahina SEMAR-BITAH 
• Mohamed BEN ELLEFI 
• Aljaz KOSMERLJ 
• Peter CLARKE 
• Tu...
Problem Definition 
• Exhibition data (time, date etc.) is stored in natural text 
in web pages and currently not stored a...
User workflow
Technology stack
British Museum schema
RETRIEVING BRITISH MUSEUM DATA 
• Example SPARQL query used to retrieve labels that will 
map to Dbpedia concept: 
• In th...
Event Registry Integration 
http://eventregistry.org/json/event?action=getEvents&conceptUri=http://en.wikipedia.org/wiki/V...
Title 
Location 
Descriptio 
n 
Time??? 
Event Registry Integration
Event Registry Integration 
eventregistry.org/json/event?action=getEvent&eventUri=443353&resultType=dateMentionAggr 
{"dat...
Application PROTOTYPE
lESSONS LEARNED 
• Important to use URIs rather than text for presenting 
content (DBpedia URI vs “China exhibition”) 
• I...
ASSUMPTIONS 
• Assumes a very, very general search over the British 
Museum data. 
• Assumes a method of aggregating poten...
Future Work 
• Integrate databases from a broader range of museums, 
i.e. worldwide. 
• Extend functionality by creating u...
THANK YOU! 
EventRegistry 
DBpedia 
SPARQL
Exhibition recommendation using British Museum data and Event Registry - ESWC SSchool 14 - Student project
Upcoming SlideShare
Loading in …5
×

Exhibition recommendation using British Museum data and Event Registry - ESWC SSchool 14 - Student project

643 views

Published on

Exhibition recommendation using British Museum data and Event Registry - ESWC SSchool 14 - Student project

Published in: Science
  • Be the first to comment

  • Be the first to like this

Exhibition recommendation using British Museum data and Event Registry - ESWC SSchool 14 - Student project

  1. 1. Exhibition recommendation using British Museum data and Event Registry
  2. 2. Introduction • Group 1: Ninja Turtles • Kahina SEMAR-BITAH • Mohamed BEN ELLEFI • Aljaz KOSMERLJ • Peter CLARKE • Tutors • Barry NORTON • José María GARCÍA • •
  3. 3. Problem Definition • Exhibition data (time, date etc.) is stored in natural text in web pages and currently not stored as Linked Data. • Why is this an issue? • What could we do with this information in RDF format?
  4. 4. User workflow
  5. 5. Technology stack
  6. 6. British Museum schema
  7. 7. RETRIEVING BRITISH MUSEUM DATA • Example SPARQL query used to retrieve labels that will map to Dbpedia concept: • In this case, British Museum prefLabels correspond to DBpedia concepts with high accuracy.
  8. 8. Event Registry Integration http://eventregistry.org/json/event?action=getEvents&conceptUri=http://en.wikipedia.org/wiki/Vikings&conce ptUri=http://en.wikipedia.org/wiki/Museum&conceptUri=http://en.wikipedia.org/wiki/Art_exhibition&category Uri=http://www.dmoz.org/Reference/Museums&sortBy=rel&conceptLang=eng&page=0&count=25&resultTy pe=list
  9. 9. Title Location Descriptio n Time??? Event Registry Integration
  10. 10. Event Registry Integration eventregistry.org/json/event?action=getEvent&eventUri=443353&resultType=dateMentionAggr {"dateMentionAggr":[{"date":"2001-09-11", "count":1.000000}, {"date":"2014-03-05", "count":5.000000}, {"date":"2014-02-27", "count":1.000000}, {"date":"2014-06-22", "count":1.000000}]} Filter out according to some predefined time window M I N M A X Exhibition Duration 2014-02-27 2014-06-22
  11. 11. Application PROTOTYPE
  12. 12. lESSONS LEARNED • Important to use URIs rather than text for presenting content (DBpedia URI vs “China exhibition”) • In order to be more useful we would need to scrape prefLabels from a number of instances. • Time durations are inaccurate.
  13. 13. ASSUMPTIONS • Assumes a very, very general search over the British Museum data. • Assumes a method of aggregating potentially large number of EventRegistry results of a unreliable quality.
  14. 14. Future Work • Integrate databases from a broader range of museums, i.e. worldwide. • Extend functionality by creating user profiles, e.g. student studying Egyptian Civilization. • More accurate time period ranges. • Utilize tools such as SILK or LIMES for interlinking of museum concepts with DBpedia concepts. • Create more robust and feature-rich interface.
  15. 15. THANK YOU! EventRegistry DBpedia SPARQL

×