Active Ontology ActOn: An semantic Information Integration approach for dynamic distributed informat

1
Active Ontology (ActOn) An semantic Information
Integration approach for dynamic distributed
information sources

• Wei Xing, Oscar Corcho, Carole Goble, Marios
  Dikaiakos
• Information Management Group
• School of Computer Science
• University of Manchester

2
Outline

• Motivation
• Introduction ActOn
• Overview of ActOn architecture
• Prototype implementation
• ActOn-based information service deployment on
  EGEE
• Experiment evaluation
• Conclusions and future work

3
Motivation

• Providing Just enough, Just in Time information
  for a large-scale, dynamic distributed system.
• Workflow management
• Meta-scheduling
• Resource discovery
• Service level agreement
• Problems with the existing information services
• Expensive
• In-Correct
• Meaningful
• Complex Query

4
Challenges

• Information about most of Grid entities needs to
  be updated very frequently
• The latency of retrieving a piece of information
  may be longer than its change frequency
• Information of a Grid entity consists of multiple
  attributes, whose values can be normally obtained
  from different geographically-distributed
  information sources
• Heterogeneous information sources are
  available/unavailable dynamically.

5
Proposed solution

• Active Ontology (ActOn).
• an ontology-based information integration
  approach that can be used to generate and
  maintain up-to-date metadata for a dynamic,
  large-scale distributed system.
• Graphic data model
• Metadata cache
• Intelligent semantic metadata management
• Lifetime-control
• Dynamic information source selection
• Updating-on-demand

6
Related work (1)

• Existing Grid Information Services
• MDS
• Monitoring-based information service
• Expensive
• Ladp data model (weak in query)
• RGMA
• Relational database servlets
• Poor performance in a large scale, dynamic
  distributed environment
• rigorous with change (database schema)
• BDII
• MDS-based information service
• Cached information in each service (local or
  regional)
• High performance

7
Comparison
8
Related Work (2)

• Limitations of the traditional ontology-based
  information integration approach
• Not prepared for highly dynamic information
• Not fault-tolerant and robust to changes in the
  information source availability
• Examples
• TSIMMS, Manifold, InfoSleuth.

9
Comparison
10
Introduction of the ActOn

• An Ontology-based Meta-aggregation Information
  Service
• Domain Ontologies Grid Ontology
• Service-oriented within S-OGSA architecture
• Main functionalities
• Integrate diverse data sources
• Extremely dynamic (appear, temporarily
  unavailable, disappear)in grid environment
• Generate and maintain up-to-date metadata
• Dynamic, large-scale distributed systems
• Provide a uniform interface
• Accessing and discovery of metadata

11
Overview of ActOn Architecture
12
System Components

• ActOn is comprised of
• software components
• Metadata Scheduler (Msch)
• Information Source Selector (ISS)
• Metadata Cache Service (McS)
• A set of Wrappers
• Knowledge components (Ontologies)
• Domain ontologies
• Information source ontologies

13
Metadata Scheduler (Msch)

• It is designed to apply an update-on-demand
  policy to cache metadata.
• That is, the cached metadata is not updated until
  it is stale when being queried, so as to avoid
  unnecessary updates.
• We adopt event-driven mechanisms to cope with
  that policy.

14
Defined events

• We have defined three types of events that can
  trigger the update process,
• Application-specific events
• Need specific information updating at certain
  time.
• Query events
• Query when it is out lifetime
• System-related events
• Related events happen, which may cause change.

15
Information Source Selector (ISS)

• The Information Source Selector (ISS) is used to
  find the most suitable information source from
  the set of available sources, which are described
  as instances of the Information Source Ontology.
  Information sources can be any system (database,
  file, service, etc.) that contains relevant
  information.

16
Information source store

• Define the class of each kind of information
  source in information source ontology, and its
  property accordingly.
• Generate the instances of information sources
  into the information source store.
• Query the store for a information source for a
  piece of the specific information.
• Select the information source.

17
Metadata Cache Service (McS)

• The Metadata Cache (MC) stores and manages the
  metadata obtained from the information sources,
  together with its timestamp and lifetime
  information, so that it can check whether such
  property values are still valid or not (e.g.,
  lifetime control) when it receives a query event
  that involves them.

18
Lifetime control for the semantic metadata

• Control the life time of metadata in property
  level
• No fixed lifetime
• No same lifetime for a multiple-property metadata
• Property life time control
• House-keeping service
• Managing the lifetime of the metadata

19
The house-keeping service
20
Wrappers

• The query translator
• General query by GUI
• Translating to a particular query language
• Ontology-based
• Schema mapping
• Define the access api and access point (instances)

21
Knowledge components

• Domain ontology
• Define the concepts and properties of a specific
  doamin
• In our example grid ontology
• Information source ontology
• Define the concepts and properties of each kind
  of information sources
• In our example EGEE information source ontology

22
Overview of the ontologies
23
Example 1 Grid domain ontology

• The Core Grid ontology

24
Example 2 the information source ontology

• EGEE information source ontology

25
Prototype Implementation

• Java application Jena RDF API

26
ActOn Deployment on EGEE

• EGEE production testbed includes 200 sites, 35K
  CPUs, 13 Petabyes of storage and runs 30K-70K
  jobs per day on behalf of 100 VOs
• The User Interface that we used to access the
  EGEE Grid is the UI (ui.tier2.hep.manchester.ac.uk
  ) at University of Manchester

27
Experiment evaluation

• Objectives of the experiment evaluation
• A fair systematic approach to measure
  Information quality of different Grid
  information services
• Evaluate information quality of two EGEE
  information services (BDII and RGMA) and ActOn.

28
Experiment set-up

• Selected two metrics commonly used in information
  retrieval
• Precision the proportion of relevant information
  retrieved, out of all the information retrieved.
• Recall the proportion of relevant information
  that is retrieved, out of all the relevant
  information available.
• EGEE tesbed
• Three Grid nodes Manchester, UCY, and Belgrade
• A set of Unix shell scripts

29
Queries used in Evaluation

• Query 1 Find all the Computing Elements (CEs)
  that support the BIOMED Virtual Organisation
  (VO).
• Query 2 Find all the CEs that support the BIOMED
  VO and have more than 100 CPUs available.
• Query 3 Find all the CEs that support the MPI
  running environment.
• Query 4 Find all the CEs that support the BIOMED
  VO, have more than 100 CPUs available, and
  support the MPI running environment.
• Query 5 Find all the CEs where GATE (Geant4
  Application for Tomographic Emission) can be run.
• Query 6 Find all the CEs that support the BIOMED
  VO, have more than 100 CPUs available, and where
  GATE can be run.

30
Query Translation
31
Results of Query 1 in different ISs
32
Results
33
Remarks

• BDII has weak query capabilities
• RGMA is not able to relate information available
  in different tables
• RGMA is very sensitive to the registering and
  availability of information providers at a given
  point in time
• Some complex queries cannot be answered by one
  type of information service in isolation

34
Conclusions and future work

• Active Ontology (ActOn)
• an ontology-based information integration
  approach,
• ActOn overcomes some of the limitations
• information quality
• availability and robustness
• response time
• Future work
• Information quality validation
• Algorithm for dynamic lifetime predication

35
ActOn publications

• W. Xing, O. Corcho, C. Goble, and M.
  D. Dikaiakos, "A Grid Information Service based
  on an Intelligent Information  Integration
  Architecture." European Semantic Web Conference
  2007 (ESWC-2007), poster, June, 2007.
• W. Xing, O. Corcho, C. Goble, and M. D.
  Dikaiakos, "Active Ontology An Information
  Integration Approach for Dynamic Information
  Sources", in the 2nd International Workshop on
  Semantic and Grid Computing (SGC-07),  in
  conjunction with the International Symposium on
  Parallel and Distributed  Processing and
  Applications (ISPA-2007). (accepted for
  publication).
• W. Xing, O. Corcho, C. Goble, and M. D.
  Dikaiakos, "Information Quality Evaluation for
  Grid Information Services." CoreGRID Symposium
  2007. In conjunction with EuroPar 2007 (accepted
  for publication).
• W. Xing, O. Corcho, C. Goble, and M. D.
  Dikaiakos, "ActOn A Semantic  Information 
  Service  for  EGEE",  the 8th IEEE/ACM
  International Conference on Grid Computing (Grid
  2007) (accepted for publication).(22 accept
  rate)

36

• Thank you for your attention!
• Comments and questions!

37
(No Transcript)
38
(No Transcript)