Presentation is loading. Please wait.

Presentation is loading. Please wait.

Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding1 Ontological Reasoning for Natural Language Understanding Fabian M. Suchanek.

Similar presentations


Presentation on theme: "Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding1 Ontological Reasoning for Natural Language Understanding Fabian M. Suchanek."— Presentation transcript:

1 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding1 Ontological Reasoning for Natural Language Understanding Fabian M. Suchanek Supervisors: Prof. Dr-Ing. Gerhard Weikum Dr. habil. Peter Baumgartner Colloquium

2 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding2 Overview 1. Introduction to the Problem 2. Our Approach 1. SUMO-Ontology 2. Text Analysis 3. KRHyper 3. Conclusions and Outlook

3 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding3 Text Entailment Hypothesis: „Elvis is still alive!“ Text:...Yes, Elvis is alive......Elvis is not alive......Elvis comes to Saarbrücken on January 1st......Elvis is living incognito in a small Russian village...

4 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding4 Plan of Attack KRHyper-Rules subclass(funeral, socialinteraction). subclass(literacyattribute, traitattribute). subclass(mixedlanguage, spokenhumanlanguage). subclass(pidginlanguage, spokenhumanlanguage). subclass(creolelanguage, spokenhumanlanguage). subclass(human,animal). subclass(deafsignlanguage, manualhumanlanguage).). subclass(radio, electricdevice).). subclass(contractcarrier, transportationcompany). subclass(commoncarrier, transportationcompany). subclass(terminalbuilding, building). subclass(trainstation, transitterminal). subclass(representationalartwork, artwork). subclass(artwork, artifact). subclass(singlefamilyresidence, permanentresidence). subclass(hotel, commercialagent). subclass(hotel, temporaryresidence). subclass(hotel, residentialbuilding). subclass(residentialbuilding, residence). subclass(residentialbuilding, building). subclass(temporaryresidence, residence). subclass(permanentresidence, residence). subclass(residence, stationaryartifact). subclass(room, stationaryartifact). subclass(building, stationaryartifact). comesTo(elvis,saarbruecken). subclass(watercraft, vehicle). subclass(train, railvehicle). subclass(wagon, landvehicle). subclass(railvehicle, landvehicle). subclass(motorcycle, roadvehicle). subclass(artificialsatellite, satellite). subclass(communicationdevice, engineeringcomponent). subclass(contractcarrier, transportationcompany). SUMO- Ontology Text

5 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding5 SUMO-Ontology – Concept Hierarchy Entity PhysicalAbstract ConstrictorSnake Anaconda Human

6 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding6 instance(P, intentionalProcess)  agent(P, A) => holdsDuring(when(P), attribute(A, awake)) SUMO-Ontology – Axioms

7 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding7 instance(P, intentionalProcess)  agent(P, A) => holdsDuring(when(P), attribute(A, awake)) SUMO-Ontology – Axioms Functions

8 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding8 instance(P, intentionalProcess)  agent(P, A) => holdsDuring(when(P), attribute(A, awake)) SUMO-Ontology – Axioms Functions Nested Formulae O

9 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding9 SUMO-Ontology – Functions alive(wife(elvis)). equal(wife(elvis), priscilla). alive(priscilla). alive(X)  equal(X,Y) => alive(Y) blueeyed(X)  equal(X,Y) => blueeyed(Y) blond(X)  equal(X,Y) => blond(Y) naive(X)  equal(X,Y) => naive(Y) cute(X) /\ equal(X,Y) => cute(Y) slim(X) /\ equal(X,Y) => slim(Y) simple(X) /\ equal(X,Y) => simple(Y) blueeyed(X) /\ equal(X,Y) => blueeyed(Y) ?

10 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding10 SUMO-Ontology – Functions alive(wife(elvis)). equal(wife(elvis), priscilla). alive(priscilla). equal(wife(elvis), X) => alive(X) equal(wife(elvis), priscilla). Correct  Complete  0 additional clauses 

11 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding11 instance(P, intentionalProcess)  agent(P, A) => holdsDuring(when(P), attribute(A, awake)) SUMO-Ontology – Axioms Functions  Nested Formulae

12 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding12 SUMO-Ontology – Nested Formulae thinks(bob, alive(priscilla)  alive(elvis)). alive(elvis)  alive(priscilla) alive(priscilla)  alive(elvis) thinks(bob, alive(elvis)  alive(priscilla) ).

13 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding13 SUMO-Ontology – Nested Formulae thinks(bob, false  beautiful(priscilla)). alive(elvis)  ¬alive(elvis) false false  beautiful(priscilla) ? thinks(bob, alive(elvis)  ¬alive(elvis) ). O

14 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding14 Propositional Equivalence p(t 1,… a  b,… t n )  p(t 1,… a,… t n )  p(t 1,… b,… t n ) p(t 1,…  x s,… t n )   x p(t 1,… s,… t n ) p(t 1,… not(a),… t n )  not(p(t 1,… a,… t n ))

15 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding15 Propositional Equivalence p(t 1,… q(t' 1,…t' n' ),… t n ) ~~~~~>  x p(t 1,…,x,… t n )  q(t' 1,…t' n',x) p(t 1,… not(q(t' 1,…t' n' )),… t n ) ~~~~~>  x p(t 1,…,x,… t n )  q(t' 1,…t' n',neg(x)) says(bob,a) says(bob,not(b)) ab

16 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding16 Propositional Equivalence, Example thinks(bob, alive(elvis)  ¬alive(elvis) ) => thinks(bob, beautiful(priscilla) ) thinks(bob, wife(elvis,priscilla)) wife(X,Y) husband(Y,X) => thinks(bob, husband(priscilla,elvis)) knows(bob, alive(elvis)) knows(X,Y) => believes(X,Y) => believes(bob, alive(elvis))

17 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding17 Plan of Attack KRHyper-Rules subclass(funeral, socialinteraction). subclass(literacyattribute, traitattribute). subclass(mixedlanguage, spokenhumanlanguage). subclass(pidginlanguage, spokenhumanlanguage). subclass(creolelanguage, spokenhumanlanguage). subclass(human, animal). subclass(deafsignlanguage, manualhumanlanguage).). subclass(radio, electricdevice).). subclass(contractcarrier, transportationcompany). subclass(commoncarrier, transportationcompany). subclass(terminalbuilding, building). subclass(trainstation, transitterminal). subclass(representationalartwork, artwork). subclass(artwork, artifact). subclass(singlefamilyresidence, permanentresidence). subclass(hotel, commercialagent). subclass(hotel, temporaryresidence). subclass(hotel, residentialbuilding). subclass(residentialbuilding, residence). subclass(residentialbuilding, building). subclass(temporaryresidence, residence). subclass(permanentresidence, residence). subclass(residence, stationaryartifact). subclass(room, stationaryartifact). subclass(building, stationaryartifact). comesTo(elvis, Saarbruecken). subclass(watercraft, vehicle). subclass(train, railvehicle). subclass(wagon, landvehicle). subclass(railvehicle, landvehicle). subclass(motorcycle, roadvehicle). subclass(artificialsatellite, satellite). subclass(communicationdevice, engineeringcomponent). subclass(contractcarrier, transportationcompany). SUMO- Ontology Text

18 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding18 Text Analysis Elvis comes to Saarbrücken Noun Verb Prep Noun Comp VP S O

19 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding19 Text Analysis Elvis comes to Saarbrücken Noun Verb Prep Noun Comp VP S type=propertense=pres number=sg type=proper number=sg LFG-Analysis

20 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding20 Mapping a Sentence to SUMO Elvis comes to Saarbrücken ?

21 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding21 Mapping a Sentence to SUMO Elvis comes to Saarbrücken BodyMotion

22 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding22 Mapping a Sentence to SUMO Elvis comes to Saarbrücken BodyMotion ? ? Agent Destination O

23 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding23 Mapping a Sentence to SUMO Elvis comes to Saarbrücken BodyMotion ? ? Agent Destination dsubj

24 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding24 Mapping a Sentence to SUMO Elvis comes to Saarbrücken BodyMotion ? Agent Destination to+location

25 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding25 Mapping a Sentence to SUMO Elvis comes to Saarbrücken BodyMotion Agent Destination instance(b,bodymotion). agent(b,elvis). destination(b,saarbruecken).

26 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding26 Plan of Attack KRHyper-Rules subclass(funeral, socialinteraction). subclass(literacyattribute, traitattribute). subclass(mixedlanguage, spokenhumanlanguage). subclass(pidginlanguage, spokenhumanlanguage). subclass(creolelanguage, spokenhumanlanguage). subclass(human, animal). subclass(deafsignlanguage, manualhumanlanguage).). subclass(radio, electricdevice).). subclass(contractcarrier, transportationcompany). subclass(commoncarrier, transportationcompany). subclass(terminalbuilding, building). subclass(trainstation, transitterminal). subclass(representationalartwork, artwork). subclass(artwork, artifact). subclass(singlefamilyresidence, permanentresidence). subclass(hotel, commercialagent). subclass(hotel, temporaryresidence). subclass(hotel, residentialbuilding). subclass(residentialbuilding, residence). subclass(residentialbuilding, building). subclass(temporaryresidence, residence). subclass(permanentresidence, residence). subclass(residence, stationaryartifact). subclass(room, stationaryartifact). subclass(building, stationaryartifact). comesTo(elvis, Saarbruecken). subclass(watercraft, vehicle). subclass(train, railvehicle). subclass(wagon, landvehicle). subclass(railvehicle, landvehicle). subclass(motorcycle, roadvehicle). subclass(artificialsatellite, satellite). subclass(communicationdevice, engineeringcomponent). subclass(contractcarrier, transportationcompany). SUMO- Ontology Text

27 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding27 KRHyper ر First-Order-Reasoning System ر Developed at the University of Koblenz/Germany by P. Baumgartner and C. Wernhard ر Given some facts and rules, calculates a set of entailed facts Facts+ Rules entailed O

28 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding28 Disjunctive Logic Programs Rule: H 1  …  H m ← B 1  …  B n p(a) ← p(b). {} {p(a)} {p(b)} {p(a),p(b)} Model for a ground rule H 1  …  H m ← B 1  …  B n : A set M of literals such that (  j,1  j  n: B j  M) =>  i, 1  i  m: H i  M p(X)  q(X) ← r(X,a)  s(X,b)

29 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding29 KRHyper Tableaux Algorithm q(a). p(X) ← q(X). r(X)  s(X) ← p(X). false ← r(a). q(a) p(a) r(a) false s(a) {q(a),p(a),s(a)}

30 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding30 attribute(A, alive) ← agent(P,A)  instance(P,process). subclass(bodyMotion, process). instance(b,bodyMotion). agent(b,elvis). Application SUMO- Ontology Text attribute(elvis,alive). instance(b,process). Hypothesis false ← attribute(elvis,alive).

31 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding31 Results ر It is possible to enhance question answering by a kind of common sense reasoning ر There is still a long way to go to practical applicability O

32 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding32 Problems 1. Analysis-related problems 2. SUMO-related problems 3. Content-related problems “Coal stocks got a lift“ “The report said…“ Hi! A woman hit her partner.  ? A man was hit. ♥♥

33 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding33 Outlook Idea: Learn the Ontology from the Web Web (Wikipedia) Free Text + HTML + Ontological Knowledge

34 Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding34 Conclusion References: KRHyper: http://www.uni-koblenz.de/~wernhard/krhyper SUMO:http://www.ontologyportal.org Comp. Linguistics:http://www.coli.uni-saarland.de Thesis:http://www.mpii.mpg.de/~suchanek/personal Elvis is alive, but there‘s still much work to do.


Download ppt "Fabian M. Suchanek Ontological Reasoning for Natural Language Understanding1 Ontological Reasoning for Natural Language Understanding Fabian M. Suchanek."

Similar presentations


Ads by Google