36
A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations Andre Freitas, Kartik Asooja, Joao B. Jares, Stefan Decker, Ratnesh Sahay SWAT4LS 2014

A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Embed Size (px)

Citation preview

Page 1: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

A Semantic Web Platform for

Automating the Interpretation of

Finite Element Bio-simulations

Andre Freitas, Kartik Asooja, Joao B. Jares,

Stefan Decker, Ratnesh Sahay

SWAT4LS 2014

Page 2: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Insight Centre for Data Analytics

Page 3: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Goals

Automate the interpretation of finite element (FE) biosimulations ...

... by providing a supporting a symbolic representation of FE data.

Insight Centre for Data Analytics Slide 3

Page 4: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Reproducibility of FE Simulations

Insight Centre for Data Analytics Slide 4

Page 5: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Efficiency & Automation of FE Simulations

Insight Centre for Data Analytics Slide 5

Page 6: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Motivational Scenario: Cochlear

mechanics

Insight Centre for Data Analytics Slide 6

Page 7: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Characteristics of the FE Domain

• Relatively small set of concepts

Insight Centre for Data Analytics Slide 7

Page 8: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Characteristics of the FE Domain

• But difficult to represent • Physics, geometrical models, topological relations, algoithmic,

mathematics

Insight Centre for Data Analytics Slide 8

Page 9: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Characteristics of the FE Domain

• Most data is at the numeric level

• Highly dependent on visualization (man in the middle)

Insight Centre for Data Analytics Slide 9

Page 10: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Dimensions of a FE Bio-simulation

Insight Centre for Data Analytics Slide 10

Page 11: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Geometrical Model

Insight Centre for Data Analytics Slide 11

Page 12: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Physics Model

• FE equilibrium for solid

• FE equilibrium for fluid

Insight Centre for Data Analytics Slide 12

Page 13: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Numerical Models/Solvers

• Incremental-iterative implicit solution scheme

Insight Centre for Data Analytics Slide 13

Page 14: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Experimental Data

• A

Insight Centre for Data Analytics Slide 14

Page 15: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

And others ... (which are not covered here)

• Anatomical

• Physiological

• Histological

• ...

02 May 2014 Insight Centre for Data Analytics Slide 15

Page 16: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Lid-driven cavity flow

Insight Centre for Data Analytics Slide 16

Physical Model

Solver

FEM Model

If there a vortex close to

the lid?

Page 17: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Lid-driven cavity flow

Insight Centre for Data Analytics Slide 17

Physical Model

Solver

FEM Model

If there a vortex close to

the lid?

Informal definition of a

valid simulation

Page 18: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Numerical Data Interpretation

02 May 2014 Insight Centre for Data Analytics Slide 18

informal description of

the simulation

Rules using references

to anatomical, physical

and data feature

elements

Is translated into

Multiple simulations

Feature extraction

Interpretation = rules

applied over data at

the symbolic level

Page 19: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Automatic Interpretation

Insight Centre for Data Analytics Slide 19

Expected physical behavior (Experiment intent):

Velocity in X starts at zero at the bottom of the box

followed by a slow velocity decrease reaching a

minima which is followed by a very fast velocity

increase close to the lid.

Numeric Level

Symbolic Lifting

IF

Predicates

Page 20: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data View

Insight Centre for Data Analytics Slide 20

Data Selection

y

0.05

Page 21: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Feature Extraction (Symbolic Lifting)

Insight Centre for Data Analytics Slide 21

Minima=(0.055,-0.20)

fast increase

slow decrease

followed by

(avg first derivative > 35)

velocity starts

at 0 at the

bottom

maximum

velocity is 0.93

at the lid

Based on the TEDDY

ontology

Page 22: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Interpretation Statements

Insight Centre for Data Analytics Slide 22

:DataView1 :hasDimensionY :VelocityX .

:DataView1 :hasDimensionX :DistanceFromTheCavityBase .

:DataView1 :x0 “0.0"^^xsd:double .

:DataView1 :y0 “0.0"^^xsd:double .

:DataView1 :hasMinimumX “-0.055"^^xsd:double .

:DataView1 :hasMinimumY “-0.20"^^xsd:double .

:DataView1 :hasFeature :PositiveSecondDerivative .

:DataView1 :hasBehaviour :BehaviourRegion1 .

:DataView1 :hasBehaviour :BehaviourRegion2 .

:BehaviourRegion1 :avgFirstDerivative “-3.63"^^xsd:double .

:BehaviourRegion1 :hasFeature EndRegion .

:BehaviourRegion1 :hasFeature :Decreases .

:BehaviourRegion1 :hasFeature :DecreasesSlowly .

:BehaviourRegion2 :avgFirstDerivative “33.35"^^xsd:double .

:BehaviourRegion2 :hasFeature EndRegion .

:BehaviourRegion2 :hasFeature :Increases .

:BehaviourRegion2 :hasFeature :IncreasesFast .

:BehaviourRegion1 :isFollowedBy :BehaviourRegion1 .

: LidSimulation :hasInterpretation :ValidVelocityBehaviour .

Data Analysis Rule

Page 23: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Rules

Insight Centre for Data Analytics Slide 23

CONSTRUCT

{ :LidSimulation sif: hasInterpretation :ValidVelocityBehaviour }

WHERE {

?dataview rdf:type dao:DataView .

?dataview dao:hasFeature ?x .

...

}

IF( minima(velocity) is negative AND

decreases very slowly(velocity) AND

increases very fast (velocity) )

VALID VELOCITY BEHAVIOUR

SPARQL Rule

Page 24: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Workflow

Insight Centre for Data Analytics Slide 24

Page 25: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Workflow

Insight Centre for Data Analytics Slide 25

Page 26: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Workflow

Insight Centre for Data Analytics Slide 26

Page 27: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Workflow

Insight Centre for Data Analytics Slide 27

Page 28: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Data Analysis Workflow

Insight Centre for Data Analytics Slide 28

Page 29: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Conceptual Model Excerpt

Insight Centre for Data Analytics Slide 29

Page 30: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Conceptual Model Excerpt

Insight Centre for Data Analytics Slide 30

Page 31: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Going back to the Cochlea

simulation scenario

02 May 2014 Insight Centre for Data Analytics Slide 31

Page 32: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Output Data Views

Insight Centre for Data Analytics Slide 32

Page 33: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Feature Extraction

Insight Centre for Data Analytics Slide 33

:DataView1 :hasDimensionY :BasilarMembraneMagnitude .

:DataView1 :hasDimensionX :DistanceFromTheCochleaBasis .

:DataView1 :hasFeature :isSingleWave .

:DataView1 :hasMaximumAmplitude “0.0031 "^^xsd:double.

:DataView1 :hasMaximumY “0.0020 e^-6 "^^xsd:double .

:DataView1 :hasMaximumX “14"^^xsd:double .

:DataView1 :hasMinimumY “-0.0011 e^-6 "^^xsd:double .

:DataView1 :hasMinimumX “17"^^xsd:double .

Page 34: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Demonstration (Video)

Insight Centre for Data Analytics Slide 34

http://bit.ly/1rEZYh7

Page 35: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Take-away message

• Contemporary science demands new infrastructures to

scale scientific discovery in a complex knowledge

environment.

• Numerical data is everywhere, not only in FE simulations.

• In this work we started exploring how to represent and

extract numerical data features to a conceptual level.

• Which could match user intents specified as rules in the

data.

Insight Centre for Data Analytics Slide 35

Page 36: A Semantic Web Platform for Automating the Interpretation of Finite Element Bio-simulations

Future Directions

• Better integration of the proposed representation and data

analysis framework to the TEDDY conceptual model.

• Use of the feature set and rules as a heuristic method to

prune the simulation configuration space.

Insight Centre for Data Analytics Slide 36