The Terminator's origins or how the Semantic Web could endanger Humanity.

882
-1

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
882
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
2
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

The Terminator's origins or how the Semantic Web could endanger Humanity.

  1. 1. Miel Vander Sande [@Miel_vds]TERMINATOR’S ORIGINS or how the semantic web could endanger humanity
  2. 2. Working on the Semantic Web ‘s bright future
  3. 3. is also working onthe Semantic Web‘s dark side
  4. 4. Misuse by HumansMisuse by Machines
  5. 5. Misuse by Humans
  6. 6. Content Content ContentContent A A Ageneration Automatic rule based content A Intelligent and generic agents Content High quality Linked Data A big chunk is social data A Ginormous Linked Open Data cloud
  7. 7. Force false knowledge extraction Manipulate semantic driven tasks{ ?x ex:Article ?y. ?y ex:subject dbpedia:Barack_Obama. If type == ex:FrontPage}=> #create html to present article{ ?x ex:category ex:FrontPage }.{ ?x a ex:Article .=>{ ?x rdfs:comment “<script> … open websocket … </script>” }.
  8. 8. ”One of the defining characteristics of asuccessful information system is its level of exploitation by spammers.” - Ian Davis
  9. 9. Fully context aware, easily produced and reproduced<http://.../John> a foaf:Person; foaf:firstname “John”; foaf:knows <http://.../Marc> . Hey John,<http://.../Marc> a foaf:Person ; Marc here. I saw you ex:attends <http://.../ISWC> . attend ISWC soon. Want to join me for<http://.../ISWC> a swpo:Conference ; the concert of The swpo:hasLocation <http://.../Boston> . Who: http://spam.url<http://.../Concert_TheWho> a music:Concert ; music:venue <http://.../TD Garden> .
  10. 10. Provenance manipulation fools proof and trust Data Proof & Trust WebID needs needs based on Provenancehttps://test.net/#mvs a prov:Person, prov:Agent;:post a prov:Entity, sioc:Post; sioc:title “Buy Mentos, the freshmaker"^^xsd:string; prov:wasAttributedTo <https://test.net/#mvs>; prov:generatedAtTime "2011-07-16T01:52:02Z"^^xsd:dateTime .
  11. 11. Malicious links trick automated processes Image WormPhotogallery owl:sameAs owl:sameAs owl:sameAs A Image Image
  12. 12. Misuse by Machines
  13. 13. ”Machine intelligence is Acceptance ofnot human intelligence.” machine intelligence - Chris Welty Fully deployed web of things
  14. 14. Judgement day? Today: drones, Stuxnet How much trust will we put in machines and the Semantic Web?
  15. 15. We need to proactively brainstorm about misuse investigate fail-safe approachesdiscuss "acceptable machine behaviour" before its too late? Discuss! Miel Vander Sande [@Miel_vds]
  16. 16. Attributions• Photo title slide by Ujang Uung• Photo slide #1 by Cayusa• Photo slide #2 by morteza bahmani• Photo slide #3 by Hugh Hart & ed6505• Photos slide #4 by Moyo Ogunjale & Dark Lord• Photo slide #6 by Huafeng Xie• Photo slide #14 by Universal Studios

×