Jakub J. Moskal - PowerPoint Presentation

Slide1

Jakub J. MoskalMieczyslaw “MITCH” M. KOKARBRIAN E. ULICNY November 19, 2014

Comprehension of RDF Data Using Situation Theory and Concept MapsSlide2

Outline2Lots of RDF data. Querying with SPARQL produces complicated RDF graphsObjective: Generate “simple” Concept MapsSituation Theory (Barwise

, Perry, Devlin)STO: Situation Theory OntologyProcess outline and processing stepsExamplesConclusionsSlide3

Abundance of Data3Analysts are required to sift through tremendously large amounts of dataKeyword-based queries yield poor resultsStructured data is needed

The number of RDF data sets is growing rapidlyEven though RDF data is structured, it can be very difficult to analyze

Source:

http://

lod-cloud.net

/Slide4

Linked Open Data cloud diagramAs of 08/30/20144

Linking Open Data cloud diagram 2014, by Max Schmachtenberg, Christian Bizer

, Anja Jentzsch and Richard Cyganiak. http://

lod-cloud.net

/Slide5

Example QueryQuery: What were the circumstances of Richard H. Barter’s death?RDF Data: SPARQL Endpoint: http://

dbpedia.org/sparqlSPARQL query:

PREFIX

dbpedia

-owl

: <http://

dbpedia.org

/ontology/>

DESCRIBE

?resource

WHERE

{

?

resource

dbpedia-owl:abstract

?abstract. FILTER langMatches(lang(?abstract), "EN" ). FILTER REGEX(str(?abstract), "Richard H. Barter")}LIMIT 10

5Slide6

Query Result: RDF graph6Slide7

Our Approach7Objective: Given a query, transform the input RDF graph to a Concept Map

that:Provides answer to the queryContains facts that are relevant to the query (context)Is

more abstract than the original RDF graph(easier to comprehend)Approach:

Use key aspects of Situation Theory of

Barwise

and Perry (extended and formalized by Devlin)

M

ap the

problem to this theory and implement algorithms for constructing concept maps based on such a

frameworkSlide8

Expected Benefits8Increased analyst productivityEasier comprehensionTailored visualization

Explanatory factsImproved quality of analyst productsFewer false alarmsMore detections of relevant eventsEnriched fact base via inferenceAugmented with situation types and their instances

Integration with other analyst toolsExport to standard formatsSlide9

Concept Maps9Slide10

A Bit of Situation Theory10

- Infon

- S “supports” Infon

- Situation Type

- Abstract Situation

- Definitional Query

- Inferring situations

and their types

- “Relevance” –

via entailmentSlide11

Situation Theory – Relevance ReasoningRelevant entities with respect to a given query Q are those entities that are necessary for proving that a specific set of facts SQ supported by a situation satisfies

Q.

11Slide12

Situation Theory – Why? It grounds meaning in the world, rather than in the language (unlike in FrameNets)It allows specifying views of the world (situations) that are globally inconsistent, but locally consistentSituations are first-class citizens – they have their own relations and attributesMeaning of a declarative sentence is a relation between utterances and described situations

12Slide13

STO Ontology13Slide14

CONOPS14Slide15

Representing Queries(in terms of ElementaryInfon and Situation Type)15

“Did an insurgent visit a weapons cache?”Expressible in pure OWL:InsurgentWeaponsCacheSituation

≡ Situation and (supportedInfon

some

(

ElementaryInfon

and

(

anchor1

some

Insurgent)

and

(

anchor2

some

WeaponsCache

)

and (relation value visit))) “Which insurgents spied on a relative?”Not expressible in pure OWL, requires use of variablesRules are necessary, for instance:Situation(s) ∧ ElementaryInfon(i) ∧ Object(a1) ∧ Object(a2) ∧ Relation(spiedOn) ∧ supportedInfon(s, i) ∧ anchor1(i, a1) ∧ anchor2(i, a2) ∧ relation(i, spiedOn) ∧ Insurgent(a1) ∧ Person(a2) ∧ relative(anchor1, anchor2) → RelativeSpySituation(s) Slide16

Answering Queries: Process16Slide17

A Running Example (based on SynCOIN)17Query:

Which known insurgents are connected to people who have been to a weapons cache?WCSit ≡ Situation and

(supportedInfon some (ElementaryInfon

and

(anchor1

some

Insurgent)

and

(anchor2

some

(

Person and

hasBeeonTo

some

WeaponsCache

)))

and

(relation

value isConnectedTo))) Initial facts:Slide18

(1) Domain Inference18Infer implicit facts about the domainIf necessary, add additional axioms to the dataset

We added a few axioms to SynCOIN:Slide19

(2) Situation Reasoning19Analyze situation type definitions (both in OWL and rules)Extract

relevant relations used in definitions (visit and spiedOn in previous examples). Then extract

relevant individuals.For each relation rel that is part of a situation type:

For

each pair of individuals

a

1

and

a

2

that are

associated

with each other by the property

rel

:

Assert

that there is an individual

s

of RDF type

sto:Situation Assert that there is an individual i of RDF type sto:ElementaryInfon, supported by situation s Assert the following facts: (i anchor1 a1), (i anchor2 a2) and (i relation rel) Slide20

Example – cont.

20

Initial Graph:

Current Answer:Slide21

(3) Context Derivation21Derive the context for the answerFind

relevant facts: all individuals and relations that are relevant to the situation that represents the answer to the queryDerivation based on domain-independent rules, which backtrack OWL inferenceCurrently: Property chain, sub-property, transitive propertyExample derivation rule for transitive property:

For a situation s, and a query q, if

s

satisfies the query:

For

every fact (

i

1

rel

i

2

) relevant to

s

and an individual

i

3

, if

rel is a transitive property and if (i1 rel i3) and (i3 rel i2) are facts asserted in the knowledge base: Add (i1 rel i3) and (i3 rel i2) as facts relevant to s. Slide22

Example – cont.

22

Previous Step:

Current AnswerSlide23

(4) Simplification23Context derivation is likely to produce a lot of “noise”W

e need to remove facts that are relevant to a situation, but that are not necessary to comprehend the graph Simplification based on domain-independent rulesExample simplification rule for sub-property relation between relevant relations:For a situation

s, and a query q, if s

satisfies the query

:

For every relation

r

1

and

r

2

relevant

to

s

, if

r

1

is a

sub-property of r2:For every two facts (i1 r1 i2) and (i1 r2 i2) that are both relevant to s: Remove (i1 r2 i2) from the context of s. Slide24

Example – cont.

24

Previous step:

Final answer:Slide25

Conclusions25Objective: simplify answers to queries against RDF dataApproach: use Situation Theory (Barwise

, Perry, Devlin)Approximate Situation Theory formalization by using STO: Situation Theory Ontology, OWL and RulesQueries represented by STO:ElementaryInfon and STO:SituationUsed OWL axioms to enhance reasoning about the domainDeveloped domain-agnostic rules for inferring relevant situations, situation types, relations and individuals in situations

Developed context derivation rules Developed context simplification rules Developed a prototype and showed (on examples) that it works

BaseVISor

was used for inference

To make it practical, “meta-reasoning” was needed.Slide26

Future Work26More domain-independent inference rules neededClusteringInference-driven

generalizationMachine LearningFeedback collected from GUIConcept/Link removal (affects transformation rules)Graphical arrangement (affects clustering)Scalability

Very large scale graph databasesIntegration with data analyticsEvaluate with analysts!Slide27

Thank You