The NoRDF Project Fabian Suchanek Amazing! This talk is free of - - PowerPoint PPT Presentation
The NoRDF Project Fabian Suchanek Amazing! This talk is free of the Corona virus! (about the speaker, we dont know...) Knowledge Bases P erson subclassOf Singer type born 1935 For us, a knowledge base (KB) is a graph, where the
Fabian Suchanek Amazing! This talk is free
(about the speaker, we don’t know...)
For us, a knowledge base (KB) is a graph, where the nodes are en ti ties and the edges are relations. 2
type born 1935 Singer P erson subclassOf
(We do not distinguish T
Apple Siri 3
IBM Watson Discovered 6 kineasis proteins that relate to cancer How long was the Thirty Y ears’ War? Amazon Echo These applications feed from knowledge bases.
Plus industrial projects at Sponsored message: New version of YAGO at h ttp://yago-knowledge.org .
From YAGO
Essen tially binary facts (“triples”) in the knowledge format “RDF”: 5
In February 1998, Andrew Wake field published a paper in the medical journal The Lance t, which reported on twelve children wi th developmen tal disorders. The paren ts were said to have linked the start
ting con troversy became the biggest science story of 2002. As a resul t, vaccination rates dropped sharply. In 2011, the BMJ de tailed how Wake field had faked some of the data behind the 1998 Lance t article. Belie fs Claims Even ts Reasons Stories Falsifications ...none of which is in a knowledge base! 6
If we wan t tomorrow’s in telligen t applications to be really in telligen t, we have to extend their knowledge bases by 7 1) We have to be able to extract complex knowledge from text (“IE”) 2) We have to be able to represen t such knowledge and to reason on i t Belie fs Claims Even ts Reasons Stories Falsifications
8 Several cool approaches can extract non‐binary in formation:
arser
t spanners
Andrew Wake field published in The Lance t in 1998. Publication_even t author venue time
9 “Wake field published a paper that reported on children. Their paren ts were said to have linked the start of behavioral symptoms to vaccination. The resul ting con troversy caused vaccination rates to fall. ...” Publication RateChange Wake field paper Claim symptoms children vaccination Link paren ts vaccinationRate
direction author pub. con ten t about
by
10 Publication RateChange Wake field paper Claim symptoms children vaccination Link paren ts vaccinationRate
direction author pub. con ten t about
by
Y
that can do (part of) i t? Please le t me know! T ype here: _ _ _ _ _ _ _ _ _ _ _ _ Cross‐sen tence analysis, advanced co‐re ference resolution, standardized types of frames, relationships be tween even ts, negation, hypothe tical stances, storylines, ...
11 RateChange vaccinationRate
direction
As knowledge represen tation:
Publication Wake field paper
caused author pub.
12 As knowledge represen tation:
tifiers
For reasoning:
text logics
tation
f revision
Cannot represen t
ts believe that the company delivers a good service”
t happened because the public learned of a fraudulen t activi ty by the company”
aul says, P aul says Mary believes ” ... or if they can, they are undecidable
13 1) a very simple logic inside a con text 2) a very simple logic about con texts => a moderately simple logic in combination First‐order logic wi thout ? OWL EL? Datalog? Horn Rules? Datalog? Y
t me know! (?) (?) V agueness, fuzziness, and probabili ty: orthogonal topics
14
understand an article about a con troversial topic, allow reasoning (who said what when and why, what is the evidence, ...)
extract con troversy or belie fs wi th reasons and supporters, for companies or their products
troversies: de tect a con troversial topic on the Web (in blogs, forums, T wi tter), extract opinions, and model differen t views
>more
Understanding the argumen ts of the other side is a prerequisi te for re futing them.
15
tially fraudulen t activi ty: De tect claims that con tradict knowledge, or violate rules.
Model sequences of actions, causal relationships, and suggestions.
Allow dialogues that go beyond single-shot questions.
Analyze a law, a regulation, or a con tract, and derive what is permi tted and what is obligatory for which party.
16 Our project “NoRDF” aims to extract and model complex in formation from natural language text. We are supported by the French National Research Agency, T élécom P aris, and 4 sponsors:
17 Fabian Suchanek Professor at T élécom P aris Knowledge Bases, Reasoning, NLP Chloé Clavel Professor at T élécom P aris Affective Computing, Sen timen t Analysis We hired
aris (CNAM) as a postdoc
t
ehran) as PhD studen t
18 We are hiring PhD studen ts, postdocs, and engineers, for the project
th NLP , knowledge bases, and reasoning! Join our team! h ttps:// suchanek.name -> NoRDF
20 RateChange vaccinationRate
direction
As knowledge represen tation:
Publication Wake field paper
caused author pub.
21 RateChange vaccinationRate
direction
As knowledge represen tation:
Publication Wake field paper
caused author pub.
great, but do not allow for reasoning
tist”
aul says & P aul says X, then Mary believes X”
22 RateChange vaccinationRate
direction
For reasoning:
Publication Wake field paper
caused author pub.
23 RateChange vaccinationRate
direction
For reasoning:
Publication Wake field paper
caused author pub.
great, but do not allow for statemen ts about statemen ts
24 RateChange vaccinationRate
Annotated Knowledge Represen tations:
tifiers
Publication Wake field paper
caused author pub.
25 RateChange vaccinationRate
Annotated Knowledge Represen tations:
tifiers
Publication Wake field paper
caused author pub.
cannot deal wi th hypothe tical statemen ts cannot do reasoning
26 Big logic machinery:
text logics
27 Big logic machinery:
text logics
ts believe that the company delivers a good service”
t happened because the public learned of a fraudulen t activi ty by the company”
(or if they can, they are proposi tional logics or undecidable)
cannot quan tify over con texts Formal argumen tation has monoli thic proposi tions. Belie f revision has monoli thic agen ts. Provenance and annotated logics cannot make claims about annotations. V agueness, fuzziness, and probabili ty are orthogonal topics.