modelling data-intensive web sites with ontoweaver

27
Modelling Data-Intensive Web Sites with OntoWeaver Knowledge Media Institute The Open University Yuangui Lei, Enrico Motta, John Domingue {y.lei, e.motta, j.b.domingue}@open.ac.uk

Upload: vevay

Post on 31-Jan-2016

17 views

Category:

Documents


0 download

DESCRIPTION

Modelling Data-Intensive Web Sites with OntoWeaver. Knowledge Media Institute The Open University Yuangui Lei, Enrico Motta, John Domingue {y.lei, e.motta, j.b.domingue}@open.ac.uk. Outline. Motivation OntoWeaver Rationales Site View Modelling Presentation Modelling - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Modelling Data-Intensive Web Sites with OntoWeaver

Modelling Data-Intensive Web Sites with OntoWeaver

Knowledge Media InstituteThe Open University

Yuangui Lei, Enrico Motta, John Domingue{y.lei, e.motta, j.b.domingue}@open.ac.uk

Page 2: Modelling Data-Intensive Web Sites with OntoWeaver

Outline

• Motivation• OntoWeaver Rationales• Site View Modelling• Presentation Modelling• Customization Framework• An Example• Advantages and Limitations• Conclusions

Page 3: Modelling Data-Intensive Web Sites with OntoWeaver

Motivation

• Data-intensive web sites– Time-consuming design process– Few re-usable components – Costly maintenance process

• Solutions– Model-based web modelling methodologies

• HDM, RMM, WebML, etc.

Page 4: Modelling Data-Intensive Web Sites with OntoWeaver

Customization Layer

User ProfilesUser Model

Customization Requirements

Presentation Layer

Presentation Styles Layouts

Site View Layer

Data ElementsNavigational Elements

KA Forms Query Forms

Data Layer

Databases

Data Services

Domain Model

Current Research Practices

Page 5: Modelling Data-Intensive Web Sites with OntoWeaver

Open Issues

• Limited support for the specification of site views;– RMM, OOHDM, HDM-Lite– ARANEUS, WebML, OntoWebber, HERA

• Limited support for the specification of presentation styles and layouts;– Style sheets– Layout design

• Limited support for customization at design time as well as at run time;– The customization specification is intertwined with the site

view layer– The aspects of customization are limited

Page 6: Modelling Data-Intensive Web Sites with OntoWeaver

OntoWeaver: an Ontology-Based Approach

• Goal:– to enable the process of designing ontology-based web

applications to be carried out at the conceptual level• Rationales

– Employing the notion of ontology as the back-bone• Site View Ontology• Presentation Ontology• Customization Framework

– All aspects of a web site expressed declaratively• Advantages

– Low-cost route to data-driven web applications– All aspects of a web site amenable to reasoning

• Site Design Critiquing• Customization

Page 7: Modelling Data-Intensive Web Sites with OntoWeaver

Customization Layer

User ProfilesUser Model

Customization Requirements

Presentation Layer

Presentation Styles Layouts

Site View Layer

Data ElementsNavigational Elements

KA Forms Query Forms

Data Layer

Databases

Data Services

Domain Model

Domain Ontology

Site View Ontology

Presentation Ontology

CustomizationFramework

Page 8: Modelling Data-Intensive Web Sites with OntoWeaver

OntoWeaver: an Ontology-Based Approach

• Goal:– to enable the process of designing ontology-based web

applications to be carried out at the conceptual level• Rationales

– Employing the notion of ontology as the back-bone• Site View Ontology• Presentation Ontology• Customization Framework

– All aspects of a web site expressed declaratively• Advantages

– Low-cost route to data-driven web applications– All aspects of a web site amenable to reasoning

• Site Design Critiquing• Customization

Page 9: Modelling Data-Intensive Web Sites with OntoWeaver

OntoWeaver Architecture

OWWeb Site

InferenceEngine

Site Builder

Presentation Specification

Site View Specification

Site Designer

Site View & Presentation Ontologies

User Profiles

Domain Ontology

Ontology Editor

Customization Rule Base

Site CustomizerOnline Page

Builder

User Ontology

CustomizedView

Page 10: Modelling Data-Intensive Web Sites with OntoWeaver

Modelling the Site View Layer

hasSubComponent

hasSubResource

Site

Resource

Component

hasIndexResource

hasComponent

SubResource

MetaData

hasMetaData

hasResource

LinkItem

hasOutput

Output Input

CommandhasInput

hasCommand

hasLinkItem

Page 11: Modelling Data-Intensive Web Sites with OntoWeaver

Examples of the Site View Constructs

KA component

Input element

Command ElementOutput

Elements

Page 12: Modelling Data-Intensive Web Sites with OntoWeaver

Data Component

Dynamic Output

Page 13: Modelling Data-Intensive Web Sites with OntoWeaver

Site View Constructs Summary

• Composite User Interface Constructs– Site, SiteResource, ResourceComponent

• Component Primitives:– DataComponent– KAComponent– SearchComponent

• Pre-defined services

• Atomic User Interface Constructs– Output, DynamicOutput, Command, Input

• Navigational Constructs– LinkItem, DynamicLinkItem, Parameter

Page 14: Modelling Data-Intensive Web Sites with OntoWeaver

Modelling the Presentation Layer

• Presentation Styles– Fonts, colours, etc– Template-based approach

• Templates: generic templates, widget templates, data component templates

• Presentation objects: attaching templates with site view components

• Layouts– Organization of site view components– Layout modelling

• Text Layout• Component Layout

• Presentation Model– SitePresentation

Page 15: Modelling Data-Intensive Web Sites with OntoWeaver

Presentation Model

SitePresentation

TemplatehasPresentation

hasTemplatehasLayout

Presentation

Layout

PresentationTemplate

ComponentURI

hasTemplatehasComponentURI AreaLayout

hasAreaLayout

ComponentURI

Page 16: Modelling Data-Intensive Web Sites with OntoWeaver

Customization Framework

• Declarative Site Model– Site view specifications;– Site presentation specifications; – Site ontologies.

• User Ontology– A model to enable the recording of user profiles– Only very basic ontology is provided, typically this will be

specialised for particular applications• User Profiles

– Information about end users in terms of user ontology.• Customization Rules

– Conditions– Customization Actions

• Jess Inference Engine

Page 17: Modelling Data-Intensive Web Sites with OntoWeaver

OntoWeaver Architecture

OWWeb Site

InferenceEngine

Site Builder

Presentation Specification

Site View Specification

Site Designer

Site View & Presentation Ontologies

User Profiles

Domain Ontology

Ontology Editor

Customization Rule Base

Site CustomizerOnline Page

Builder

User Ontology

CustomizedView

Page 18: Modelling Data-Intensive Web Sites with OntoWeaver

Example: Building a Paper Review System

• Four types of users• Authors

– They browse conference information– They submit papers

• Reviewers – They get the papers assigned for review from the web site;– They submit their reviews to the web site;

• Program committee members– They browse review recommendations.

• Chair– Has access to everything, notifies authors, etc.

Page 19: Modelling Data-Intensive Web Sites with OntoWeaver

Design Process

• Domain Ontology Design:• Site View Design:

– Designing navigational structures– Composing user interfaces for web pages

• Presentation Design– Specifying visual appearance for user interface elements– Specify layout for user interface elements

• Customization Design:– Identifying potential audiences;– Creating different site models for different user groups;– Defining customization rules for individual users or groups;

Page 20: Modelling Data-Intensive Web Sites with OntoWeaver

OntoWeaver Site Designer

Page 21: Modelling Data-Intensive Web Sites with OntoWeaver

A General View

Page 22: Modelling Data-Intensive Web Sites with OntoWeaver

User Group Specified Customization

Page 23: Modelling Data-Intensive Web Sites with OntoWeaver

Rule-based Customization

Page 24: Modelling Data-Intensive Web Sites with OntoWeaver

Reviewers get extra links for obtaining assigned papers and submitting reviews;

Two corresponding new web pages have been created;

Customized web pages are generated on the fly when users whose role is “Reviewer” login onto the paper review system and make requests about web pages.

A Customized View For Paper Reviewers

Page 25: Modelling Data-Intensive Web Sites with OntoWeaver

Advantages

• High level specification– The design can be made much simpler

• Automatic site generation through ontology mapping

• Customization support– The modular approach– The customization framework

• Site maintenance– Tools support for manual and semi-automatic

maintenance

Page 26: Modelling Data-Intensive Web Sites with OntoWeaver

Limitations

• Focusing on data presentation, data acquisition, and data querying

• Remote web services are not considered• Semantics of web pages are not explored

Page 27: Modelling Data-Intensive Web Sites with OntoWeaver

Conclusions

• OntoWeaver uses an ontology-driven approach to the design and development of data-intensive web sites.

• It provides a set of ontologies to support the modelling and specification of all aspects of data-intensive web applications.

• It provides a customization framework to allow the specification of customization requirements at a high level of abstraction.

• It provides a tool suite to support the design and development of data-intensive web istes.