University of the West of England Home Page Peter Hale My Home Page SEEDS Site Map Site Map

Text Only Site Map Text Only Site Map

User Driven Programming

PhD Poster

Peter Hale

Home page http://www.cems.uwe.ac.uk/~phale/

Introduction

SEEDS (Systems Engineering Estimation and Decision Support) team - involved in cost modelling solutions, particularly for aerospace.

Working with Rolls-Royce aerospace and Airbus.

This presentation demonstrates how a modelling Ontology can be used to automatically produce representations.

End User Programming

Why do we need to make it easier for end users to program.

Figure 1 - End User Programming

Based on data from US bureau of Labour Statistics.

Sources - http://www.cs.cmu.edu/~bam/papers/EUPchi2006overviewColor.pdf - Myers et al.

Scaffidi, C., Shaw, M., Myers, B. (2005). Estimating the Numbers of End Users and End User Programmers, IEEE Symposium on Visual Languages and Human-Centric Computing, (VL/HCC'05): 207-214 Dallas, Texas.

User Driven Programming

Figure 2 - User Driven Programming

Software development is time consuming and error prone because of the need to learn computer languages.

Mitigating this allows users to devote full effort to the problem to be solved.

User Driven Programming creates software that enables people to program with visual representation of a tree diagram.

The tree representation is translated into computer languages.

Applied to aerospace engineering but should be applicable to any subject.

A Semantic editor, Protégé, is used to create the Ontology for the cost models.

The approach is illustrated in Figure 3.

Figure 3 - Source to Result Tree Translation

Translation

The Ontology representation is translated into a computer model.

An Ontology defines relationships between things.

Relationships can be conveyed to a software model that evaluates them.

To achieve this the translator requires -

Search trigger(s) resulting from user actions.

Knowledge of the relationships between nodes in the tree.

Ability to read a equations held in a standardised mathematical form.

Rules of syntax for the language of the code to be output.

Figure 4 - Vanguard Studio Representation and Calculation

Visualisation

Figure 5 shows how the program visualises information for the spar and its' part definition, material, manufacturing processes etc.

Figure 5 - Vanguard Studio

Visualisation

This is an interactive view of the Vanguard Studio model that was read from Protégé.

Changing the figures and pressing recalc results in a new calculation.

This tree is translated into SVG (Scalable Vector Graphics) and JavaScript for an interactive CAD style view.

Translations from Tree Based View to Component Diagram

Figure 5 shows the spar translated to XML and visualised using Flash multimedia. The spar is then translated and visualised in an SVG based interactive diagram.

Figure 5 - Tree View - XML - Flash - SVG Representation

Others in this kind of research

Kurt Cagle -Understanding XML - http://www.understandingxml.com/

General Electric - ACUITy enterprise modelling tool -Jena Conference Paper - http://jena.hpl.hp.com/juc2006/proceedings/crapo/paper.pdf

 - An Ontology-Based Architecture for Adaptive Work-Centered User Interface Technology - A Aragones, J Bruno, A Crapo, M Garbias.

Jena Conference Proceedings - http://jena.hpl.hp.com/juc2006/proceedings.html

Orbeon - http://www.orbeon.com/ - Orbeon XForms Presentation Server.

Protégé - Conference and Project Information - http://protege.stanford.edu/community/conferences.html

UWE - Christophe Bru - http://www.cems.uwe.ac.uk/~cbru/

Vanguard Global Knowledge Portal - http://wiki.vanguardsw.com/

Chris Wallace - http://www.cems.uwe.ac.uk/~cjwallac/

Developed by Peter Hale, 2005. The University of the West of England, Bristol, UK.