Knowledge Mobilization: Architectures, Models and Applications Juan Gómez Romero Doctoral Thesis July 2008 Advisor: Miguel Delgado Calvo-Flores Department.

Slides:



Advertisements
Similar presentations
Ontology-Based Computing Kenneth Baclawski Northeastern University and Jarg.
Advertisements

Andrea Maurino Web Service Design Methodology Batini, De Paoli, Maurino, Grega, Comerio WP2-WP3 Roma 24/11/2005.
2 Introduction A central issue in supporting interoperability is achieving type compatibility. Type compatibility allows (a) entities developed by various.
ARCHITECTURES FOR ARTIFICIAL INTELLIGENCE SYSTEMS
® IBM Software Group © 2006 IBM Corporation Rational Software France Object-Oriented Analysis and Design with UML2 and Rational Software Modeler 04. Other.
ICT 1 “Putting Context in Context: The Role and Design of Context Management in a Mobility and Adaptation Enabling Middleware” Marius Mikalsen Research.
Technical Architectures
Context Representation and Reasoning with Formal Ontologies Juan Gómez-Romero 1,2, University Carlos III of Madrid (Spain) Fernando Bobillo 2, University.
Semantic description of service behavior and automatic composition of services Oussama Kassem Zein Yvon Kermarrec ENST Bretagne France.
Towards Ubiquitous Government Services through Adaptations with Context and Views in a Three-Tier Architecture Dan Hong, SC Cheung, SMIEEE Department of.
April 15, 2005Department of Computer Science, BYU Agent-Oriented Software Engineering Muhammed Al-Muhammed Brigham Young University Supported in part by.
Marakas: Decision Support Systems, 2nd Edition © 2003, Prentice-Hall Chapter Chapter 1: Introduction to Decision Support Systems Decision Support.
Marakas: Decision Support Systems, 2nd Edition © 2003, Prentice-Hall Chapter Chapter 7: Expert Systems and Artificial Intelligence Decision Support.
ReQuest (Validating Semantic Searches) Norman Piedade de Noronha 16 th July, 2004.
Distributed Collaborations Using Network Mobile Agents Anand Tripathi, Tanvir Ahmed, Vineet Kakani and Shremattie Jaman Department of computer science.
What is adaptive web technology?  There is an increasingly large demand for software systems which are able to operate effectively in dynamic environments.
Software Issues Derived from Dr. Fawcett’s Slides Phil Pratt-Szeliga Fall 2009.
Course Instructor: Aisha Azeem
Community Manager A Dynamic Collaboration Solution on Heterogeneous Environment Hyeonsook Kim  2006 CUS. All rights reserved.
The chapter will address the following questions:
Semantic Web Technologies Lecture # 2 Faculty of Computer Science, IBA.
An Intelligent Broker Architecture for Context-Aware Systems A PhD. Dissertation Proposal in Computer Science at the University of Maryland Baltimore County.
Knowledge Management in Geodise Geodise Knowledge Management Team Liming Chen, Barry Tao, Colin Puleston, Paul Smart University of Southampton University.
Advances in Technology and CRIS Nikos Houssos National Documentation Centre / National Hellenic Research Foundation, Greece euroCRIS Task Group Leader.
Chapter 6 System Engineering - Computer-based system - System engineering process - “Business process” engineering - Product engineering (Source: Pressman,
Ontology Development Kenneth Baclawski Northeastern University Harvard Medical School.
TOWARDS INTEROPERABILITY IN TRACKING SYSTEMS: AN ONTOLOGY-BASED APPROACH Juan Gómez Romero Miguel A. Patricio Jesús García José M. Molina Applied A.I.
TC Methodology Massimo Cossentino (Italian National Research Council) Radovan Cervenka (Whitestein Technologies)
Mihir Daptardar Software Engineering 577b Center for Systems and Software Engineering (CSSE) Viterbi School of Engineering 1.
Database System Concepts and Architecture
Architecture Tutorial 1 Overview of Today’s Talks Provenance Data Structures Recording and Querying Provenance –Break (30 minutes) Distribution and Scalability.
A service-oriented middleware for building context-aware services Center for E-Business Technology Seoul National University Seoul, Korea Tao Gu, Hung.
Odyssey A Reuse Environment based on Domain Models Prepared By: Mahmud Gabareen Eliad Cohen.
Ontology Summit 2015 Track C Report-back Summit Synthesis Session 1, 19 Feb 2015.
© DATAMAT S.p.A. – Giuseppe Avellino, Stefano Beco, Barbara Cantalupo, Andrea Cavallini A Semantic Workflow Authoring Tool for Programming Grids.
11 CORE Architecture Mauro Bruno, Monica Scannapieco, Carlo Vaccari, Giulia Vaste Antonino Virgillito, Diego Zardetto (Istat)
Chapter 3 DECISION SUPPORT SYSTEMS CONCEPTS, METHODOLOGIES, AND TECHNOLOGIES: AN OVERVIEW Study sub-sections: , 3.12(p )
Page 1 WWRF Briefing WG2-br2 · Kellerer/Arbanowski · · 03/2005 · WWRF13, Korea Stefan Arbanowski, Olaf Droegehorn, Wolfgang.
1 Introduction to Software Engineering Lecture 1.
A Context Model based on Ontological Languages: a Proposal for Information Visualization School of Informatics Castilla-La Mancha University Ramón Hervás.
Using Several Ontologies for Describing Audio-Visual Documents: A Case Study in the Medical Domain Sunday 29 th of May, 2005 Antoine Isaac 1 & Raphaël.
Christoph F. Eick University of Houston Organization 1. What are Ontologies? 2. What are they good for? 3. Ontologies and.
PICTURE your design. Purpose : Functions & Content Functions the facilities that make the content of the ICT useful for relevant users and other ICT’s.
11 CORE Architecture Mauro Bruno, Monica Scannapieco, Carlo Vaccari, Giulia Vaste Antonino Virgillito, Diego Zardetto (Istat)
Ontology Mapping in Pervasive Computing Environment C.Y. Kong, C.L. Wang, F.C.M. Lau The University of Hong Kong.
Volgograd State Technical University Applied Computational Linguistic Society Undergraduate and post-graduate scientific researches under the direction.
Issues in Ontology-based Information integration By Zhan Cui, Dean Jones and Paul O’Brien.
Digital Libraries1 David Rashty. Digital Libraries2 “A library is an arsenal of liberty” Anonymous.
Architecture View Models A model is a complete, simplified description of a system from a particular perspective or viewpoint. There is no single view.
An Ontology-based Approach to Context Modeling and Reasoning in Pervasive Computing Dejene Ejigu, Marian Scuturici, Lionel Brunie Laboratoire INSA de Lyon,
Providing web services to mobile users: The architecture design of an m-service portal Minder Chen - Dongsong Zhang - Lina Zhou Presented by: Juan M. Cubillos.
Approach to building ontologies A high-level view Chris Wroe.
A Portrait of the Semantic Web in Action Jeff Heflin and James Hendler IEEE Intelligent Systems December 6, 2010 Hyewon Lim.
Copyright 2007, Information Builders. Slide 1 iWay Web Services and WebFOCUS Consumption Michael Florkowski Information Builders.
Rule Engine for executing and deploying the SAGE-based Guidelines Jeong Ah Kim', Sun Tae Kim 2 ' Computer Education Department, Kwandong University, KOREA.
Introduction: Databases and Database Systems Lecture # 1 June 19,2012 National University of Computer and Emerging Sciences.
XML and Distributed Applications By Quddus Chong Presentation for CS551 – Fall 2001.
Informatics for Scientific Data Bio-informatics and Medical Informatics Week 9 Lecture notes INF 380E: Perspectives on Information.
Managing Data Resources File Organization and databases for business information systems.
Defects of UML Yang Yichuan. For the Presentation Something you know Instead of lots of new stuff. Cases Instead of Concepts. Methodology instead of the.
Semantic Graph Mining for Biomedical Network Analysis: A Case Study in Traditional Chinese Medicine Tong Yu HCLS
UNIT 1.
Chapter 2 Database System Concepts and Architecture
Object-Oriented Software Engineering Using UML, Patterns, and Java,
OPM/S: Semantic Engineering of Web Services
Software Design and Architecture
Copyright © 2011 Pearson Education, Inc. Publishing as Pearson Addison-Wesley Chapter 2 Database System Concepts and Architecture.
MANAGING KNOWLEDGE FOR THE DIGITAL FIRM
Systems Analysis and Design in a Changing World, 6th Edition
THREE TIER MOBILE COMPUTING ARCHITECTURE
Presentation transcript:

Knowledge Mobilization: Architectures, Models and Applications Juan Gómez Romero Doctoral Thesis July 2008 Advisor: Miguel Delgado Calvo-Flores Department of Computer Science and Artificial Intelligence University of Granada

Overview We have investigated solutions to the problems of building Knowledge-Based Systems that deliver knowledge obtained from large information sources to nomadic users. 07/17/20082Knowledge Mobilization: Architectures, Models and Applications

Overview Mobile devices and communication networks have given rise to a shift from desktop applications to mobile systems. Mobile or nomadic systems can be accessed from anywhere at anytime by using mobile technologies. Intelligent systems can take advantage of mobile technologies and innovative functionalities can be implemented, but problems arise. We aim at providing solutions to the problems that appear in systems that deliver ellaborated knowledge to nomadic users. Contributions can be applied in non-mobile systems. 07/17/20083Knowledge Mobilization: Architectures, Models and Applications

1. Introduction The problem Knowledge-Based System (KBS): Software system that manages represented knowledge to solve complex decision problems. KBSs provide support for decision-making by supplying the right person with the right information at the right time. But nowadays… –the right information has to be obtained by integrating distributed and heterogeneous information sources. –the right person can be located at anywhere. –the right time can be any moment. 07/17/2008 4Knowledge Mobilization: Architectures, Models and Applications Mobile Technologies Knowledge Representation

1. Introduction The problem Use of mobile technologies in KBSs poses several challenges: –Technological issues. Mobile networks and devices have limited capabilities: screen size, bandwidth, etc. –Computational issues. Mobile systems have intrinsic features that make them more complex than a simple extension of classical systems: Delivery of knowledge to distributed and sparse users (nomadic). Adaptation to the context of the user (context-awareness). Knowledge Mobilization (KMob) is a recent approach that tackles computational issues of mobile KBSs with the aim of improving Knowledge Management procedures. 07/17/2008 5Knowledge Mobilization: Architectures, Models and Applications

Context-aware model Knowledge Mobilization review IASO system Conclusions & future work 1.2. Methodology Design artifacts State of the art Prototype Evaluation Architecture

1. Introduction Structure of the thesis Chapter 2. Review and analysis of the state of the art in intelligent mobile systems and Knowledge Mobilization. Chapter 3. Abstract architecture to support the design of Knowledge Mobilization systems. Chapter 4. Context-aware knowledge representation model for Knowledge Mobilization. Chapter 5. Proof-of-concept system (IASO application). Chapter 6. Conclusions and future works. Bibliography 07/17/2008 7Knowledge Mobilization: Architectures, Models and Applications

outline Introduction The Knowledge Mobilization approach Architecture for Knowledge Mobilization Representation model for Knowledge Mobilization IASO: A Knowledge Mobilization application Conclusions and future work

2. Knowledge Mobilization Definition Keen & Mackintosh (2001). To make “knowledge available for real- time use in a form which is adapted to the context of use and to the needs and cognitive profile of the user”. Carlsson (2006). Four main tasks : –Creation of knowledge. Semantic Web, Ontologies and Fuzzy Logic. –Activation of latent knowledge. Multicriteria Optimization, Evolutionary Computing and Simulation. –Retrieval of hidden knowledge. Data and Text Mining and Text Summarization. –Delivery of knowledge. Multi-Agent Systems. 07/17/2008 9Knowledge Mobilization: Architectures, Models and Applications

1. Knowledge Mobilization Our proposal KMob addresses the challenge of building Knowledge Mobilization Systems, which are: –Ubiquitous. Accesible from anywhere, at anytime, using mobile technologies. –Proactive. Discover what information is needed. –Declarative. Users do not specify how information has to be obtained, but which is their situation and what information they need. –Context-aware. Behavior is adapted to context. –Integrative. Heterogeneous information sources, technologies and devices. –Concise. Summarize and tailor gathered data. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

Data: Mobile Technologies Information: Ontologies Knowledge: Fuzzy Logic, MAS, Semantic Web Applications: Healthcare Knowledge Mobilization related areas 1.3. Related areas

1. Knowledge Mobilization Use case Nomadic / Ubiquitous Healthcare. –A doctor is attending to a patient outside the hospital. –Patient’s clinical history is stored in the Hospital Information System (HIS). –The doctor uses a portable device to consult the patient’s history, in order to prescribe a treatment. –The doctor retrieves a bunch of Electronic Health Records (EHRs). –The doctor filters the results manually and grasps interesting information. –Typical scenario of Knowledge Mobilization. The mobile device can be unable to process the information obtained, or maybe the doctor has not enough time to review it (information overload) It can happen also in non-mobile systems. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

outline Introduction The Knowledge Mobilization approach Architecture for Knowledge Mobilization Representation model for Knowledge Mobilization IASO: A Knowledge Mobilization application Conclusions and future work

3. Architecture for KMob Rationale General software architectures cannot be directly extended to the Knowledge Mobilization context. Specific requirements (ubiquitous, proactive, declarative, etc.) and issues (communication, context-awareness). Contribution Meta-architecture, i.e. an abstract schema of the components, relations and operations of a Knowledge Mobilization system. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

2. Architecture for KMob AML The architecture is described with multi-agent terminology (MAS abstractions are used to describe distributed systems). Specification of the architecture with the Agent Modeling Language (AML) AML is a semi-formal visual language for specifying, modeling, and documenting systems in terms of concepts from MAS theory. Extends the UML meta-model. Advantages: well-documented, supported by visual tools, practical perspective. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

2. Architecture for KMob AML 07/17/ Knowledge Mobilization: Architectures, Models and Applications EntitiesRepresentation Agent Resource Environment Role Service Ontology Context AssociationsRepresentation Social Association Play Association Service Provision Service Usage DiagramsRepresentation SocietyGeneral view of the architecture EntityDetailed structure of an entity Protocol Sequence / Communication Specify communication acts between entities MAS DeploymentImplementation

Knowledge Mobilization: Architectures, Models and Applications 3.3. description of the architecture Knowledge provider Special service provider that manages a large knowledge base in the system, as well as incorporates other information sources (which may be external). Mobile / Nomadic requester Mobile device (cell phone, PDA), which may have very limited computational capabilities. Service provider Implement the services provided by the system: large database querying, real-time data supply, interface with knowledge bases… General components of KMob systems

Knowledge Mobilization: Architectures, Models and Applications Society diagram of the architecture (simplified) Desktop Agent Agents running on application servers Nomadic Agent Agents running on mobile devices Local Knowledge Model External Knowledge Model Services Provided / requested by the agents Roles Set of actions that an agent acquire to provide or request a service 3.3. description of the architecture

3. Architecture for KMob Frameworks The meta-architecture must be specialized for each specific problem. The meta-architecture does not state how systems should be implemented. –Which development platform should be used to implement a service which provides knowledge about patients’ clinical histories? The application designer must decide how the architecture is instantiated and which technologies are going to be used to implement it. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

3. Architecture for KMob Frameworks Three possible distributed technology frameworks: –Multi-Agent. Direct implementation with a MAS platform (JADE). Pro: Independent components that require complex coordination policies. Con: MAS platforms require a considerable amount of CPU resources. –Tuplespace. Use of shared repository of knowledge with an elemental structure (Linda, Javaspaces). Pro: Simple mechanism to achieve communication and coordination. Con: Tuplespaces require a considerable amount of network resources. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

3. Architecture for KMob Frameworks Three possible distributed technology frameworks: –Client-Server. Request-reply communication (HTTP-based, WS). Pros: –The most simple schema. –Allows the client to delegate most of the processing to the server (very limited devices can participate). Con: –Does not allow complex interaction patterns. It will be used in our application, since it satisfies our requirements. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

outline Introduction The Knowledge Mobilization approach Architecture for Knowledge Mobilization Representation model for Knowledge Mobilization IASO: A Knowledge Mobilization application Conclusions and future work

4. Representation model for KMob Rationale Knowledge Mobilization systems require a formalism to easily represent and manage knowledge. Ontologies are a representation formalism that promotes knowledge integration, sharing and reuse. Ontologies are based on Description Logics (DLs), a family of logics with well-defined semantics specially designed to represent structured knowledge. Description Logics are classified in levels (and named) according to their expressivity, which determines the complexity of reasoning with the logic. The Semantic Web uses ontologies to represent metadata and offers several tools, such as the standard OWL language. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Rationale Knowledge Mobilization formalisms are expected to solve information overload issues. To avoid information overload, only significant knowledge must be provided to users. What is significant? It depends on user circumstances: location, preferences, previous actions, etc. → Context Use of context knowledge to determine what is significant and summarize available knowledge. Knowledge Mobilization ontologies must provide support to represent, manage, and reason with context knowledge. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Rationale Contribution Meta-model, i.e. a design pattern to create context-aware ontologies that avoid information overload. Significance ontologies to represent which information of the domain is relevant in a given context. CDS (Context-Domain Significance) pattern formulated in . Directly translatable to OWL (≈  (D) ), the most expressive DL level considered. In several cases, fuzzy knowledge must be considered. Extension of the pattern using fuzzy Description Logics. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Formulation Base ontologies: –Context ontology (  C ): vocabulary to describe context situations. –Domain ontology (  D ): ontology to represent domain-specific knowledge. New significance ontology: CDS ontology (  S ) –Complex contexts ( C i ): Concepts created using terms of  C. –Complex domains ( D j ): Concepts created using terms of  D. –  -connection (  i,j or P i,j ): A concept linking a complex context C i and a complex domain D j. Denotes that D j is significant in situation C i. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Example 07/17/ Knowledge Mobilization: Architectures, Models and Applications Domain ontology Context ontology

4. Representation model for KMob Reasoning 07/17/ Knowledge Mobilization: Architectures, Models and Applications Domain knowledge I significant in a scenario E. Algorithm (implemented in the CDS API) : –Retrieve the complex contexts C n more general than E. –Retrieve the  -connections P n,m involving C n. –Retrieve the complex domains D m involved in P n,m. –Retrieve the concepts I of the domain more specific than D m. Complete and decidable Complexity is determined by C i and D j (E XP T IME -complete for  )

Knowledge Mobilization: Architectures, Models and Applications 4.5. Protégé CDS plug-in Context side Context ontology Complex contexts Domain side Domain ontology Complex domains  -connections Query tab

4. Representation model for KMob Fuzzy extension of the CDS pattern Limitations of the crisp ontology design pattern: –Imprecise knowledge cannot be represented E.g.: A patient is slightly unconscious –Partial similarities between contexts cannot be represented E.g.: Anaphylaxis is quite similar to sepsis –Relevance relations cannot hold to a degree E.g.: Blood-borne diseases are less relevant than drug intolerances Contribution Fuzzy extension of the crisp meta-model, i.e. a design pattern to create fuzzy context-aware ontologies that avoid information overload and allow vague knowledge to be managed. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Fuzzy extension of the CDS pattern The significance ontology is a fuzzy ontology (fCDS) created with an adaptation of the crisp rules of the CDS pattern. The fuzzy significance ontology is expressed with the fuzzy Description Logic f . Fuzzy DLs extends DLs to the fuzzy case (Straccia, 2006). – Concepts are fuzzy sets– Axioms hold to a degree (inclusion!) – Roles are fuzzy relations– Interpretation has fuzzy semantics Reasoning can be performed with a fuzzy DL reasoner or by reducing the fuzzy ontology to an equivalent crisp DL ontology and using a crisp inference engine (Bobillo, Delgado & Gómez- Romero). 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Fuzzy extension of the CDS pattern 07/17/ Knowledge Mobilization: Architectures, Models and Applications

4. Representation model for KMob Fuzzy extension of the CDS pattern Domain knowledge I  -significant in a scenario E. –Knowledge significant and degree of significance 07/17/ Knowledge Mobilization: Architectures, Models and Applications aggregation: min t-norm    greatest lower bound: glb = sup {  :   } Complete and decidable Complexity is determined by C i, D j, and the glbs to be calculated

outline Introduction The Knowledge Mobilization approach Architecture for Knowledge Mobilization Representation model for Knowledge Mobilization IASO: A Knowledge Mobilization application Conclusions and future work

5. IASO application Description IASO (Intelligent ASisstant for Outdoors Healthcare). KMob system to solve the Nomadic Healthcare problem for the HIS of the Hospital Clinico San Cecilio of Granada. Client-server application accesible from an intranet. The system is effective, but problems arise when: –It has to be accessed from outside the intranet. –The doctor has not enough time to review and filter patient registers to find interesting information. 07/17/ Knowledge Mobilization: Architectures, Models and Applications Representation model Architecture IASO

5. IASO application Knowledge base Three OWL ontologies have been created: Context ontology: –Based on Galen medical ontology. –Concepts ( Hemorrhage ) and relations ( galen:hasSymptom ). Domain ontology: –Created from scratch (specific for San Cecilio database). –Concepts ( Patient, Register ) and relations ( hasRegister ). Significance ontology. –The significance ontology is crisp. Since the IASO application is a verification proof of the pattern, the crisp version has been firstly used. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

5.3. IASO architecture Client agent Requires knowledge (with a mobile device) Client role Request data functions Query service Query solving service Server role Provide knowledge functions CDS Knowledge Model HIS data Server agent Provides knowledge

Knowledge Mobilization: Architectures, Models and Applications 5.4. IASO implementation SQL Bridge Links the ontological and the relational models, avoiding to import all the HIS database into the domain ontology. Implemented with D2RQ (Bizer & Seaborne, 2004).

5. IASO application Input form 07/17/ Knowledge Mobilization: Architectures, Models and Applications Query vocabulary Patient description vocabulary In-construction query Partial (conjunctive) query Patient ID Patient identification (name)

5. IASO application Output form 07/17/ Knowledge Mobilization: Architectures, Models and Applications Results Relevant registers and contents Further information Register relevant to a more specific situation that may be considered

outline Introduction The Knowledge Mobilization approach Architecture for Knowledge Mobilization Representation model for Knowledge Mobilization IASO: A Knowledge Mobilization application Conclusions and future work

6. Conclusions and future work Summary and conclusions Overall objective: –Provide integral solutions for the challenges that arise when developing mobile systems for the delivery of knowledge retrieved from large information sources to nomadic users. Operational objectives: –Distribution of knowledge in mobile systems. –Solving of information overload by summarization of available data. 07/17/ Knowledge Mobilization: Architectures, Models and Applications Architecture for Knowledge Mobilization Context-aware (fuzzy) representation model IASO application

6. Conclusions and future work Future work Future work: –Apply proposals in other problems and areas (new fields of study and domains of application!) –Architecture: Specify in detail orchestration and choreography. Introduce Semantic Web Services to describe service features. –Representation model: Compare with other Logics (non-monotonic logics). Further studies on the fuzzy extension: simplification and better support. –IASO system: Reliable deployment. Support security. Extend supporting ontologies, particularly to the fuzzy case. 07/17/ Knowledge Mobilization: Architectures, Models and Applications

end Knowledge Mobilization: Architectures, Models and Applications Juan Gómez Romero gracias thank you