Ontology-based Information Extraction from Tourism Web ... - Neries

FAW



Inst. für Anwendungsorientierte Wissensverarbeitung

Earthquake Engineering

Workshop in
eScience

Applications for Seismology

March 7
-
9 2011, Edinburgh

On finding Links between

Information Systems
and Knowledge
Based Systems

in Civil Engineering
and Seismology / Earthquake
Engineering

a.Univ.
-
Prof. Dr. Josef Küng

2

Facts and Figures

FAW

About the Institute


History

-

1990 founded as a research institute

-

1991 first year in
Hagenberg

-

1997 regularly institute of JKU

-

2005 foundation of FAW
-
GmbH


-

2005 EU
-
FP6
-
Project SAFEPIPES

-

2008 EU
-
FP7
-
Project IRIS

-

2010 EU
-
FP7
-
Project NERA



Team
(FAW
-
Institut
)

-

currently 15 persons in research and development



R&D

-

more than 100 successful finished projects and co
-
operations

-

among others currently we are coordinating (together with Dr. Wenzel, VCE)


the large EU
-
FP7 project IRIS (Integrated European Industrial Risk Reduction System)

(c) FAW
–

Johannes Kepler Universität

|

Information and Knowledge

3

Information

FAW

Current Research Domains


Information Modeling


Adaptive modeling tool


Modeling dynamic aspects of processes



Information
-
Integration


Semantic data integration (in the grid)



Datawarehouses


Loading Processes (e.g. automatic regression tests)



Information
-
Extraction


Intelligent (semantic and rule based) extraction


of structured information out of unstructured web pages




(c) FAW
–

Johannes Kepler Universität

|

4

Knowledge


Semantic Technologies, Ontologies



Using Topic Maps and Ontologies to support


queries and decisions



Ontology Enineering



Case Based Reasoning


Similarity queries in Case Based Reasoning



Application of Case Based Reasoning


Structural Health Monitoring




Application of Case Based Reasoning


in passive and active Decision Support








(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Domains

5


our famous example: tiscover
[1]

FAW

Past Research Work

Introduction



Web Based Destination
-
Management
-
System



Access to complete and up
-
to
-
date information about Tourism Holiday Destinations



Booking Functions



System Provider:

Tiscover

AG Innsbruck



Development:

FAW
-
Hagenberg

Tiscover

AG
Hagenberg




(c) FAW
–

Johannes Kepler Universität

|

6


our famous example: tiscover
[2]

FAW

Past Research Work

tiscover is more than a web page





(c) FAW
–

Johannes Kepler Universität

|

Public Terminal

(
AccessPoint
)

Reservation
&

CallCenter

Customized

Booking Engine

Internet

home/office

7


ad Information: AMMI

[1]

Meta Modeling Tool
(
A
daptive
M
odeling tool for
M
eta models and it
I
nstances)




(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work

8


ad Information: AMMI

[2]

Instance Modeling View



(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work

9


ad Information: AMMI

[3]

Administration Module




(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work

10


ad Knowledge: EU
-
Project IRIS
[1]

FAW

Current Research Work

Introduction



IRIS
–

Integrated European Industrial Risk Reduction System

–
Oct. 2008
–

Mar 2012, about 40 Partners, mainly form civil engineering domain,

4 partners from IT
-
Domain, one associated partner form Japan (University of Tokyo ) and US (Drexel University,
Stanford University)


Motivation

–
Within Current practices in risk assessment and management for industrial systems are characterized by its
methodical diversity and fragmented approaches. Integration is needed.

–
The large collaborative project IRIS is proposed to identify, quantify and mitigate existing and emerging risks to create
societal cost
-
benefits, to increase industrial safety and to reduce impact on human health and environment.



Basic Concept

–
The basic concept is to focus on diverse industrial sector’s main safety problems as well as to transform its specific
requirements into integrated and knowledge
-
based safety technologies, standards and services.



WP7: Monitoring, Assessment, Early Warning, Decision Support

–
FAW has its main task in this work package
–

setting up the decision support system.



(c) FAW
–

Johannes Kepler Universität

|

11


ad Knowledge: EU
-
Project IRIS
[2]

FAW

Current Research Work

(c) FAW
–

Johannes Kepler Universität

|

General

Structure

12

Overall Goal
–

Find the early warning point


(c) FAW
–

Johannes Kepler Universität

|


ad Knowledge: EU
-
Project IRIS
[2]

FAW

Current Research Work

13

Decision Support System



Passive Decision Support

–
Providing the right information at the right time to the decision maker

in order to support him/her.

(i.e. via Data Warehouses or via good organized (good accessible/searchable) document bases



Active Decision Support

–
A system, that uses some AI (Artificial Intelligence) methods

to elaborate a proposal to the decision maker or to do a decision autonomously.



(data mining, neural networks, support vector machines, decision trees, case based reasoning, ... )


-
> Within IRIS we work in both directions

–
Active Decision Support
-
> Case Based Reasoning

–
Passive Decision Support
-
> Semantic Networks




(c) FAW
–

Johannes Kepler Universität

|


ad Knowledge: EU
-
Project IRIS
[3]

FAW

Current Research Work

14

Active Decision Support System



Case
-
based Decision Support

(Example: Assessment of Simple Structures (Lamp Posts)



Data

–
Design (Type, Height, Material, ... )

–
Measurement (Set of selected eigenfrequencies ,

vibration measured after a stimulation)

–
Visual Inspection (Condition of post and stand,

Scratches, oxidation, condition of concrete)



Task

–
Classification of lamp post’s condition




(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work


ad Knowledge: EU
-
Project IRIS
[4]

15

Active Decision Support System



Results

–
Currently case base consists of 800 measurements of different lamp posts

–
Above 90% “correct” classifications

–
Improvement of results:

•
End
-
user can adjust parameters (attribute weights, predefined distances)
–

results are improving

•
Identify and exclude “unrepresentative cases” (where connection (parameter values


classification result) is
irreproducible)

•
In some ways the inspection process could be adapted (e.g. less “free
-
text” attributes)




In contrast to complex structures like e.g. bridges, an automated assessment of more
simple structures, as lamp posts are, looks very promising

(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work


ad Knowledge: EU
-
Project IRIS
[5]

16

Passive Decision Support System



Combining Semantic Nets and Search Engines
[1]


(Example: VCDECIS)


–
This system builds a

basic level of a wide scoped

passive Decision Support System


–
Organization/management

of an institution‘s content (documents)

to enable easier retrieval of knowledge

(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work


ad Knowledge: EU
-
Project IRIS
[6]

17

(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work


ad Knowledge: EU
-
Project IRIS
[7]

Passive Decision Support



Combining Semantic Nets and Search Engines
[2]


(Example: VCEDEIS )



Components

–
Search engine

–
Topic Map (3 layer), currently transferred to OWL

–
Web Portal

•
Document upload platform

•
Topic Map navigator incl. full
-
text search













Content
Topics

Topics

Content

18

(c) FAW
–

Johannes Kepler Universität

|

FAW

Current Research Work


ad Knowledge: EU
-
Project IRIS
[8]


Decentralized Approach


–
Each group can operate its own

Knowledge Base (KB) and

Decision Support Systems

–
IRIS Knowledge Base provides interface to partner KBs

–
Web Portal to access and administrate IRIS KB

–
Decision support (data assessment)

mainly relies on local measurement data

and on local background information (KB)

–
OWL will be the language

Knowledge Representation

(at higher level)












IRIS Ontology Landscape

IT
-
Framework, Current Big Picture

FAW

EU
-
FP7
-
Project IRIS

19

|

(c) FAW
–

Johannes Kepler Universität

CBR
-
Cycle
(Aamodt&Plaza1994):


•
Case Base
: General knowledge (knowledge
base, e.g. models, reports, rules …) and
already known cases


•
Retrieve
: Search

–
Retrieve the most similar case or cases

•
Reuse
: Adaptation

–
Reuse the information and knowledge
in that case to solve the problem

•
Revise
: Verification

–
Revise the proposed solution

•
Retain
: Learn

–
Retain the parts of this experience
likely to be useful for future problem
solving


Case Based Reasoning in General

Case Based Decision Support
[1]

FAW

EU
-
FP7
-
Project IRIS

20

|

(c) FAW
–

Johannes Kepler Universität

CBR for IRIS

•

Adopted to IRIS
-
Demands

•

More flexible


(to be used in different Domains)


Our new CBR
-
Framework for IRIS

Case Based Decision Support
[1]

FAW

EU
-
FP7
-
Project IRIS

21

|

(c) FAW
–

Johannes Kepler Universität

General Statements on Cloud Computing

Classical Computing


•
Buy & Own
:


Hardware, System Software,


Applications


(often to meet peak needs)

•
5

•
Install, Configure,

Test, Verify, Evaluate



•
Manage
:


•
. . .


•
Finally, use it


•
€€€€€

-

high Cost

Cloud Computing



Subscribe




Use










€

-

pay for what you use,



based on
QoS

(Quality of Service)

every 18 Month?

Long Term Vision ‘The IRIS Cloud’
[1]

FAW

EU
-
FP7
-
Project IRIS

22

|

(c) FAW
–

Johannes Kepler Universität

General Statements on Cloud Computing

Definition
1


A Cloud is a type of parallel and distributed system consisting of a collection of
inter
-
connected and
virtualised

computers that are
dynamically provisioned

and
presented as one or more unified computing resources based on
service
-
level
agreements

established through
negotiation

between the service provider and
consumers.


Cloud Services

•
Software as a Service
(e.g. Google Mail, … )

•
Platform as a Service
(e.g. Google App Engine, Microsoft Azure, … )

•
Infrastructure as a Service
(e.g. Amazon.com, … )


Ownership and Exposure

•
Public/Internet Clouds
(3
rd

party Cloud Infrastructure and services, available on subscription basis)

•
Private/Enterprise Clouds
(Cloud runs within a company’s data center, for internal and/or partners use)

•
Hybrid/Mixed Clouds
(mixed usage of private and public clouds)




1

Rajkumar

Buyya
, Cloud Computing and Distributed Systems
(CLOUDS) Lab, Dept. of Computer Science and Software Engineering,
The University of Melbourne, Australia


Long Term Vision ‘The IRIS Cloud’
[2]

FAW

EU
-
FP7
-
Project IRIS

23

|

(c) FAW
–

Johannes Kepler Universität

IRIS Private Cloud

Long Term Vision ‘The IRIS Cloud’
[3]

FAW

EU
-
FP7
-
Project IRIS

24

|

(c) FAW
–

Johannes Kepler Universität

IRIS Private Cloud and Mediator

Long Term Vision ‘The IRIS Cloud’
[4]

FAW

EU
-
FP7
-
Project IRIS

25

|

(c) FAW
–

Johannes Kepler Universität

IRIS Private Cloud and Consumption

Long Term Vision ‘The IRIS Cloud’
[5]

FAW

EU
-
FP7
-
Project IRIS

26

|

(c) FAW
–

Johannes Kepler Universität

•

Decision Support (WP7)



-

State: Enhanced Case Based Reasoning Framework


is in an implementation stage


Work on Active Decision Support is promising



-

Plan:

Continue on CBR, Active Decision Support


Knowledge Base and Prototypes (Proof of Concepts)


•

Data / Knowledge Integration (WP6) and


Risk Informed Design (WP8)



-

State:


IRIS System Landscape is in a stable version


Work on Integration Ontologies is ‘well on track’


(e.g. Bride Ontology is almost finished)


-

Plan:

Continue on Ontologies, keep integration in mind,


(if time, think and work more on the IRIS
-
Cloud
)

State, Plan for Next Steps

FAW

EU
-
FP7
-
Project IRIS

27

|

(c) FAW
–

Johannes Kepler Universität