User-centric Evolutionary Computation

 

 

The Web Page of the Series of Workshops at the Genetic and Evolutionary Computing Conference (GECCO) 2002 - ongoing

 

  The Domain    Links
   WORKSHOP 1- 2001    WORKSHOP 2 - 2003
   The Website (contributions)   Bibliography
   Participants   WORKSHOP 3 - 2004
   WORKSHOP 4 - 2006    WORKSHOP 5 - 2007

 

 

 

 

The Domain

There is a history of research relating to interactive evolutionary computing which, in the main, relates to partial or complete human evaluation of the fitness of solutions generated from evolutionary search. This has generally been introduced where quantitative evaluation is difficult if not impossible to achieve. Examples of application include graphic arts and animation (Sims K ,1991; Sims K.,1991b); automotive design (Graf J., Banzhaf W.,1995); food engineering (Herdy M., 1997.) and database retrieval (Shiraki H., Saito H., 1996.) Such applications rely upon a human-centred, subjective evaluation of the fitness of a particular design, image, taste etc as opposed to an evaluation developed from some analytic model.

Partial human evaluation / interaction is also in evidence. For instance, user interaction relating to an evolutionary nurse scheduling system where a schedule model provides a quantitative evaluation of a solution. However, the model may not prove adequate in terms of changing requirements, qualitative aspects etc. In this case the user must add new constraints in order to generate solutions that are fully satisfactory (Inoue T., et al., 1999). In the pharmaceutical industry Computational Biology involves the modelling of biomolecular systems. Genetic algorithms (GA) can provide the search process for the identification of optimal biomolecule combinations. The process can be enhanced, however, by the user-introduction of new combinations as an elite solution into selected GA generations (Levine D. et al, 1997). 

All the above applications utilise a major advantage of stochastic population-based search techniques. This relates to their capabilities as powerful search and exploration algorithms that can provide diverse, interesting and potentially competitive solutions to a wide range of problems. Parmee et al (1999, 2000, 2001, 2002) propose that such solutions can also provide information to the user which supports a better understanding of the problem domain whilst helping to identify best direction for future investigation. This perspective relates to human interaction when operating within ill-defined and uncertain decision-making environments in order to improve definition, increase confidence and identify innovative / creative design direction. The role here for evolutionary computation relates to exploration and the gathering of optimal information from simple conceptual models of the problem space. Such information supports model development by the user in an iterative, interactive EC environment where the first task is to evolve the problem space before attempting to solve the problem.

Top

 

 

 

Workshop 1 - July 2002.

An initial Workshop has been held at the 2002 Genetic and Evolutionary Computation Conference held at the Roosevelt Hotel, New York, July 9th-13th.  The intention of this initial Workshop was to draw together interested parties to discuss the various techniques, strategies and applications of Interactive Evolution. 

Three main areas of interaction were identified:

                     i)  Subjective selection - where interaction relates to human selection of favourite option along with, perhaps, a  simple ranking of available solutions;

                   ii) Subjective evaluation - where a more rigorous human evaluation of available solutions takes place with  a possible machine-learning component.

                    iii) Problem definition - where user interacts with evolutionary system to gather optimal information which supports the modification / reformulation of the evaluation function. Agent-based support and machine-learning possible inherent components.

This initial Workshop was attended by a diverse set of around fifteen delegates who are either active in one or more of the above areas or who wished to learn more about the various areas / techniques and to become more involved.   A broad set of examples were introduced via a series of presentations to everyone present and their particular features and potential utilities discussed.  Within the time frame available it was not possible to come to any particularly firm conclusions re the future direction of this area of research and application.  It was apparent that this first Workshop must be considered introductory and the need for further workshops at future GECCO meetings was unanimously agreed. 

Presentations

I. C. Parmee - Introduction to Workshop

I. C. Parmee - Origins of Interactive Evolutionary Computing

I. C. Parmee - Poor definition, Uncertainty and Human Factors - A Case for Interactive Evolutionary Problem Reformulation? 

P. Caleb-Solly - Interactive Evolutionary Computation - Review of Applications

 

Top

 

 

Workshop 2 - Chicago, July 2003

A second half-day Workshop was held at GECCO 2003.  The meeting comprised a diverse set of presentations addressing the three areas identified above with intermediate and final discussions.  In order to present it was not necessary to submit a paper for review as the intention was agin to identify potential areas of research and application through a broad presentation of previous work, current and future issues, possible directions and overall concepts.  Arrangements  for a special edition of Evolutionary Computation relating to Interactive Evolution are  currently underway and a Call for such an edition will be issued shortly.

Contributions to the 2003 Workshop are listed below with brief descriptions of content.   It is stressed that the Workshop represents a forum for discussion and ideas rather than an opportunity to publish detailed rigorous results. The presentations can be downloaded by clicking on the titles or by visiting referenced websites.

 

 

1)  Setting the Scene 

Ian Parmee, Advanced Computation in Design and Decision Making 

(www.ad-comtech.co.uk/ACDDM_Group.htm)

University of the West of England, UK

The presentation will briefly revisit aspects of the first Workshop highlighting relevant features and discussing the three identified areas of IEC, i.e. Subjective selection; subjective evaluation and problem definition.  A spectrum of IEC activity will be developed which will also introduce concepts of explicit and implicit interaction and explore other  forms of IEC  classification

 

2)  Exploring multiple colour spaces by an IEC technique

Tatsuo Unemi, Kayoko Mizuno

Dept. of Information Systems Science, Soka University
1-236 Tangi-machi, Hachioji, Tokyo, 192-8577 JAPAN
phone: +81-426-91-9429, fax: +81-426-91-9312
unemi@iss.soka.ac.jp http://www.intlab.soka.ac.jp/~unemi/



Through building a support tool for colour coordination by a IEC technique, we found some ideas to explore the space of colours.  This tool, named SBColor, reads SVG (Scalable Vector Graphics) data as the target drawing, sets up the genotype of enough length to represent the colours of the target in HSB (Hue, Saturation, and Brightness) color space, and provides a GUI (Graphical User Interface) for the user to breed his/her favourite co-ordination. Each component of HSB colour is represented by a floating point real number that is modified by adding some as the mutation. When the target involves many colors, it is important to facilitate another type of GUI to allow the user to restrict the
search space. We show the design of GUI to control the range of mutation for each component to realize it.

 

3)  Three levels of explicitness in interactive evolution

E. Lutton, J. Chapuis, E. Cayla, Y. Landrin-Schweitzer and Y. Semet To be  presented by Marc Schoenauer

Projet Fractales - INRIA - Rocquencourt - France

http://minimum.inria.fr/artie-fract/

Introducing three IEC projects:

Implicit interaction with an Evolutionary Algorithm takes place when ants put pheromones on the graph of an e-learning Web site proposing exercises of graded difficulty.


A more explicit form of interaction is used in a GP-based SQL-query-refinement tool that helps doctors to mine a technical database.


Finally, the artisit working with ArtieFract uses full interaction, as he not only partially sets the fitness of a Fractal image, but also directly modifies its genotype to fit his tastes.

 

Click on highlights for Powerpoint presentations in each area.

 

4) The use of interactive evolutionary design to facilitate workplace hazard
communication
.

Carnahan B.and Dorris N. Department of Industrial and Systems Engineering
Auburn University, 207 Dunstan Hall, Auburn AL. 36849

The presentation will initially comprise a DVD-based presention describing specific details of the interactive GA (representation, selection, crossover, mutation, etc.)  as a precursor to a presentation by Nathan Dorris describing controlled experiments that use the interactive GA to design and validate warning sign icons using both student and industrial populations.  Nathan is a  human factors professional with consulting experience in warning sign design and  is currently pursuing this research for his doctoral dissertation. The presentation will provide a unique perspective of the manner in which human factors specialists can use interactive evolutionary design to help promote occupational safety and health.

5)  Interacting with a Multi-Swarm

Tim Blackwell

tim.blackwell@ieee.org

 

Tim's presentation can be found on his website:  http://www.timblackwell.com


In a multi-swarm, independent swarms interact indirectly - each swarm
modifies its near environment, leaving attractors for the other swarms.
This is analogous to natural stigmergy. The evolution of the multi-swarm as
a whole can therefore be guided (i.e. by humans) by the suitable placement
of attractors. In an improvisational system all solutions found by the
population are relevant - in fact none can be discarded. An external agent
must therefore assess quality (fitness) on the fly (no pun intended), and
interact accordingly with the multi-swarm. This presentation demonstrates
the improvisational system Swarm Music, and shows how humans may
collaborate with a music-making colony of swarms.


6)  Further Developments of the Interactive Evolutionary Design System - Towards a better understanding of variable and objective space through interactive exploration

I. C. Parmee

Advanced Computation in Design and Decision-making

University of the West of England, Bristol, UK

http://www.ad-comtech.co.uk/ACDDM.htm

 

The presentation briefly revisits the Interactive Evolutionary Design System (IEDS) concept before concentrating on the analysis of multi-objective output from cluster-oriented GA results.  COGAs are being utilised for information extraction witinin the IEDS approach.  Methods for identifying mutually inclusive subsets of a number high-performance COGA regions relating to differing objectives are introduced and results of non-dominated sorting of solutions from such regions are compared to corresponding Pareto curves generated from other MOGA approaches.

 

 

 

7)  The Interactive Evolution of Configurations for "The Game"

 

P. Funes
Icosystem Corporation
Cambridge MA 02138 USA

The game can be played by people or agents. The rules are trivial: pick up two partners ("A" and "B") and apply either the "defender" rule (place yourself between A & B) or the "agressor" rule (place B btw A and yourself). The aggregate behavior is rather difficult to predict, as players keep chasing each other while each one is trying to satisfy their intrinsic goals.

Our Interactive Evolution experiment sets up a group of agents with random rules (either aggressor or defender) and partners, and lets the user evolve organized group behaviours. (You can later print "game cards" for people to play.)

So our work demonstrates the potential of IE as a way to design (find) low level rules for simple agents that create emergent, dynamic swarm behaviors. We have found rules, for example, that
create rotating spirals, jumping lines, circles, etc.

 

Top

 

 

 

Workshop 3 - Seattle, June 2004

The Workshop will again be held at the Genetic and Evolutionary Computing Conference this year to be held in Seattle, Washington, USA, June 26th - 30th 2004 (www.isgec.org/GECCO-2004) at the Red Lion Hotel.  If you are interested in presenting your work / research during the Workshop please contact Ian Parmee (ian.parmee@uwe.ac.uk) in the first instance.  It is intended, this year, to accept short papers for review and possible inclusion in the Workshop Proceedings. Please refer to the rest of the Web Site and, in particular, previous presentations for a good indication of relevant material.

Current contributions to the 2004 Workshop are listed below with brief descriptions of content.   It is stressed that the Workshop represents a forum for discussion and ideas rather than an opportunity to publish detailed rigorous results. There is still an opportunity to include further presentations in the programme. Please contact Ian Parmee if you wish to present your work.

Programme

Introduction

Ian Parmee, University of the West of England, Bristol

A brief synopsis of the material presented in the two previous Workshops and some aspects of the related discussions will presented by the Workshop Chair to set the scene for this year's meeting.

IEC as a Tool not Only for Optimizing Target Systems but Also for Observing the Human Mind

Hideyuki Takagi, Kyushu University, Japan

Interactive evolutionary computation (IEC) has been widely used to optimizetarget systems in artistic field, engineering field, and educational and entertainment filed [1]. As a target system is optimized based on a psychological measure in IEC user's mind, we may be able to observe the situation of human mind by analyzing the outputs or behavior of the optimized target system. This means that we can use IEC as a tool to analyze the situation of human mind. This is a new application approach of the IEC.

We applied the IEC to measure the human expression capability of dynamic rage of "happy - sad." We had 3 schizophrenics and 5 mental-normal students to design `happy' and `sad' impression computer graphics lighting images using IEC and asked other 33 students to evaluate the CG images using the Sheffe's method of paired comparison. Statistical test for the evaluation showed that the expression range of the three schizophrenics for
``happy-sad'' was significantly narrower than that of the normal students (p < 0.01).

This experimental result implies that IEC has potential to contribute to psychiatry, and it showed the possibility to expand the IEC applicability from conventional system optimization to new area, psychological measurement.

[1] Takagi, H. "Interactive Evolutionary Computation: Fusion  of the Capacities of EC Optimization and Human Evaluation," Proceedings of the IEEE, vol. 89, no. 9, pp. 1275-1296 (2001).

 

An IDEA for Design
 

G. Dozier, B. Carnahan, C. Seals, L.-A. Kuntz, and S.-G. Fu

Auburn University, Alabama, USA.

Evolutionary computation (EC) has been successfully applied to a wide range of design problems. There has also been an abundant amount of work in applying interactive ECs in the design of displays, robot behavior, bitmaps, etc. In the EC literature, one can also see a number of successful design applications of distributed ECs. However, to date, there has been no research in the area of interactive distributed ECs. In this paper, we present an interactive distributed evolutionary algorithm (IDEA) for the design of simple emoticons. We will discuss a variety of ways that our IDEA is currently being used including the areas of EC education, Human Factors, and User Interface Design.

Supporting Implicit Learning via the Visualisation of COGA Multi-objective Data

I. C. Parmee and J. A. R. Abraham, Advanced Computation in Design and Decision-making, CEMS      University of the West of England, Bristol, UK

The presentation speculates upon the development of human-centric evolutionary conceptual design systems that support implicit learning through the succinct visual presentation of data relating to both variable and objective space. Various perspectives of multi-objective design information support a constantly improving understanding of both subjective and quantitative relationships between variables and objectives. This information emerges from cluster-oriented genetic algorithm (COGA) output and is further defined by appropriate data mining, processing and visualization techniques. The intention is to support implicit learning and reduce complexity through the presentation of differing perspectives relating to solution / objective interaction and dependencies and continuous user interaction.. It is proposed that the developing systems could support intuitional understanding of the problem domain. 

Approaching ways to speed up the interactive evolutionary computation algorithms

Y. Saez, P. Isasi and J. C. Hernandez,
Universidad CARLOS III de Madrid, Campus de Leganes, Madrid, SPAIN.
yago.saez@uc3m.es
Javier SEGOVIA, Universidad Politecnica de Madrid, Campus de
MontePríncipe, Boadilla del Monte, Madrid, SPAIN

The talk presents a new approach in interactive evolutionary computation that helps the user in the hard task of finding an optimal solution between infinite possibilities. There are several ways of applying genetic algorithms in interactive evolutionary computation. In this talk we explain three of them in order to make an experimental comparison study. Taking as a goal (i.e.) the search for a cool logo, the Trademark Finder problem is executed with all the algorithms shown. The application made as an example is very simple in order to show clearly the results of the algorithms. The results clearly show the use of a new algorithm, based on chromosome learning heuristics for
evolving.

Identifying Relevant Symbol Design Criteria Using Interactive Evolutionary
Computation

B. Carnahan and N. Dorris,  Industrial and Systems Engineering Department,
Auburn University, Alabama, USA.

Although the computer science literature contains numerous examples that describe various interactive evolutionary computational (IEC) algorithms, few studies have focused on how to use such algorithms to elicit relevant
symbol design information from a population of IEC algorithm users. The purpose of this workshop presentation will be to illustrate how controlled experimental design involving human subject testing and multivariate statistical
analysis can be used to identify combinations of symbol design parameters that differ based on:

(1) the messages the symbols were meant to convey
(2) past personal experiences of the IEC algorithm users

Two experiments, involving both student and industrial populations, will be discussed and their findings presented.

 

Top

 

Workshop 4 - Seattle, July 2006

The Workshop will again be held at the Genetic and Evolutionary Computing Conference this year to be held  at the Renaissance Hotel in Seattle, Washington, USA, July 8th - 12th 2006 (http://www.sigevo.org/gecco-2006).  If you are interested in presenting your work / research during the Workshop please contact Ian Parmee (ian.parmee@uwe.ac.uk) in the first instance.  It is intended, this year, to accept short papers for review and possible inclusion in the Workshop Proceedings (see Call below) although, as in previous years, the Workshop is organised primarily to provide a forum for discussion.  Please refer to the rest of the Web Site and, in particular, previous presentations for a good indication of relevant material.

Call for Papers:

WORKSHOP ON USER-CENTRIC EVOLUTIONARY COMPUTATION
(Formerly the Interactive Evolutionary Computation Workshop)

to be held as part of the

2006 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE (GECCO-2006)

July 8-12, 2006 (Saturday-Wednesday)
Renaissance Seattle Hotel
Seattle, Washington, USA
Organized by ACM SIG-EVO
www.sigevo.org/GECCO-2006

SHORT PAPER (max 4 Pages) SUBMISSION DEADLINE FOR WORKSHOP: 26TH MARCH, 2006.

CHAIR: I.C. PARMEE
ACDDM LAB, BRISTOL UWE,UK.

Interactive evolutionary computing, in the main, relates to partial or complete human evaluation of the fitness of solutions generated from evolutionary search. This has been introduced where quantitative evaluation is difficult if not impossible to achieve. Examples of application include graphic arts and animation; food engineering and hazard icon design. Such applications rely upon a human-centred, subjective evaluation of the fitness of a particular design, image, taste etc as opposed to an evaluation developed from some analytic model.

Partial human interaction that complements quantitative machine-based solution evaluation is also in evidence. For instance, the user addition of new constraints in order to generate solutions that are fully satisfactory within evolutionary scheduling system or the introduction of designer-generated solutions into selected evolving generations.

Solutions can also provide information to the user which supports a better understanding of the problem domain whilst helping to identify best direction for future investigation especially when operating within poorly defined problem spaces. This supports development of the problem representation in an iterative, interactive evolutionary design and decision-making environment. Such human-centric approaches generate and succinctly present information appertaining to complex relationships between the variables, objectives and constraints that define a decision space.

In an attempt to categorise these various forms of IEC it is possible to view complete human evaluation as explicit whereas partial evaluation and interaction are less explicit, more subtle forms of human involvement. Completely implicit interaction occurs where users are unaware of their role in the evolution of a system. A simple implicit/explicit spectrum of user-centric evolutionary approaches can thus be developed.

Short papers relating to user-centric evolutionary processes across this spectrum are invited. Authors will have an opportunity during the Workshop to present the primary aspects of their research for discussion.

More details of the above with references to the examples can be found on the Workshop website (www.ad-comtech.co.uk/Workshops.htm) along with details and presentations at the 2002, 2003 and 2004 Workshops and details for the submission of short papers. The primary aim in past years has been to provide a discussive forum around a series of short presentations by participants.

 

PRESENTATIONS

The Workshop will comprise the following short presentations (15 minute duration) plus interim questions and a final, general  round-table discussion.

1.    Brief Overview of Previous IEC / UCIS Workshops

Ian Parmee, ACDDM, UWE Bristol, UK

ian.parmee@uwe.ac.uk

A brief review of the main aspects of the previous GECCO IEC / UCIS Workshops will be presented highlighting those areas that have been identified as requiring significant attention / research.

 

2.     ECSS – Evolutionary Component System

        A tool for architectural form generation and form optimization

 

Katrin Jonas, University College London

Martin Hemberg, Imperial College London and Architectural Association

katrin.jonas@ucl.ac.uk martin.hemberg@imperial.ac.uk

The presentation provides an overview of the Evolutionary Component Surface System (ECSS), a software tool for generating, analyzing and optimizing structural component surfaces. The surfaces are created from a limited set of architectural parts. We can create a large number of different surface articulations using simple rules of connectivity. The generated structures are self supporting envelopes. The form-generating process is complemented with an evolutionary algorithm to help search the space of possible outcomes. An important aspect of the tool is that design as well as structural and fabrication aspects have been considered from the outset.

3.  Co-operative OuLiPian Literature using Automatic Concept Evolution

 

Terence Fogarty and Michelle Hammond

London South Bank University 

fogarttc@lsbu.ac.uk

 

The tools and techniques employed by Interactive Evolutionary Computing (IEC) [Takagi 2001] offer procedures which put human interaction at the centre of the problem solving process.  Human based IEC systems [Korsoroff and Goldberg 2001] such as the Automatic Concept Evolver (ACE) [Fogarty 2003] have been proven to work efficiently in solving problems which involve the evolution of natural language strings. OuLiPo [Mathews and Brochi 1998] is a French literary movement founded in the 1960s which applies mathematical constraints in the creation of literature.  This paper proposes the application of the ACE methodology to arbitrate between a group of interacting authors to produce OuLiPian literature and specifies experiment to test this approach in practice.

 

4.  Interactive identification of object-oriented software design concepts using multi-objective genetic algorithms

C. L. Simons and I. C. Parmee, ACDDM, University of the West of England, Bristol, UK 

chris.simons, ian.parmee@uwe.ac.uk

The conceptual design of object-oriented software is fundamentally difficult to learn and perform, yet such conceptual design decisions have critical impact downstream in the development process. In an attempt to enable and support the human designer in this activity, a multi-objective genetic algorithm has been developed to both generate candidate conceptual design solutions derived from the problem domain, and identify candidates of superior fitness. Examples of initial results from two case studies are discussed. A number of useful and interesting design variants of high fitness are arrived at which differ from the manually produced design. It is intended that the developing system will be  highly interactive with the user both learning from and constantly adapting machine-generated solutions.  Perceived ways forward to achieve such a user-centric approach are discussed.

 

5.  Effects of User Preference on Multi-Objective Roof Truss Optimization


Breanna Bailey, Texas A&M University and Anne Raich, Lafayette College, USA.

breanna@tamu.edu,    raicha@lafayette.edu

 

The presentation details the incorporation of user design preference into an algorithm designed for the multi-objective optimization of large-span roof trusses. Several mechanisms are considered to embed preference into the optimization procedures of an implicit redundant representation (IRR) genetic algorithm.  A prediction of user-preference was included in both the selection and ranking procedures of the GA. The inclusion of user preference encourages topological exploration while preserving structurally optimal designs

 

6. An Overview of the Integration of Aesthetics, Component-based Representations and Machine-learning within a User-centric Evolutionary Design System

 

Azahar Machwe and Ian Parmee, ACDDM Lab, Bristol UWE

azahar.machwe, ian.parmee@uwe.ac.uk

This paper reviews the creation and subsequent extension of a highly user-interactive evolutionary design system with several novel features such as an inclusion of aesthetic criterion, object based representation and agent-based assembly / repair of solutions. The various components of the system and the manner in which they solve some of the problems faced by the designer during the early conceptual stages of design are described. Problems specific to  user centric EC include user-fatigue and this is particularly addressed within the system via machine-learning processes that assimilate user preferences relating to the designer's aesthetic preferenc

 

7. Using an Evolutionary Algorithm to Design Beverages

Christy Bell, Department of Chemistry, Southwestern University,
Georgetown, TX 78626
Steve Alexander, Department of Physics, Southwestern University,
Georgetown, TX 78626

We have designed an evolutionary algorithm, similar to the one designed by Todd and Latham to create art, to determine the optimum mixture of 15 liquids.   The fitness function in this project is the subjective judgement of a taster. The results of this project along with some possible implications for the other senses will be discussed.

 

 

 


 

Workshop 5 - UCL, London, July 2007

The Workshop will again be held at the Genetic and Evolutionary Computing Conference this year to be held  at UCL, London, UK , July 7th - 11th 2007 (http://www.sigevo.org/gecco-2007).  If you are interested in presenting your work / research during the Workshop please contact Ian Parmee (ian.parmee@uwe.ac.uk) in the first instance.  It is intended, this year, to accept short papers for review and possible inclusion in the Workshop Proceedings (see Call below) although, as in previous years, the Workshop is organised primarily to provide a forum for discussion.  Please refer to the rest of the Web Site and, in particular, previous presentations for a good indication of relevant material.

PRESENTATIONS

The Workshop will comprise the following short presentations (15 minute duration) plus interim questions and a final, general  round-table discussion.

1)  The Affect of User Interaction Mechanisms in Multi-objective IGA

Alexandra Brintrup

Cambridge University, Institute for Manufacturing

ab702 @ cam. ac. uk

Hideyuki Takagi

Kyushu University, Faculty of Design
takagi @ design. kyushu-u. ac. jp

2)  Supporting Free-form Design Using A Component Based Representation: An Overview

Azahar Machwe

Ian C Parmee

ACDDM Group, UWE, Bristol

azahar.machwe@uwe.ac.uk

ian.parmee@uwe.ac.uk

 

3)  Eye-Tracking Evolutionary Algorithm to minimize user’s fatigue in IEC applied to Interactive One-Max problem

Denis PALLEZ et al.

LIRIS Lab, University of Lyon

denis.pallez@unice.fr

 

4)  Implementation Issues for an Interactive Evolutionary Computation system

M.R.N. Shackelford

Advanced Computational Technologies

mshackelford@ad-comtech.co.uk

 

5)  Concept-based Multi-objective Problems and their Solution by EC

Amiram Moshaiov

Gideon Avigad

School of Mech. Eng.

Tel-Aviv University

moshaiov, gideon @eng.tau.ac

Call for Papers:

WORKSHOP ON USER-CENTRIC EVOLUTIONARY COMPUTATION
(Formerly the Interactive Evolutionary Computation Workshop)

to be held as part of the

2007 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE (GECCO-2006)

July 7-11, 2007 (Saturday-Wednesday)
University College London, UK
Organized by ACM SIG-EVO
www.sigevo.org/GECCO-2006

SHORT PAPER (max 4 Pages) SUBMISSION DEADLINE FOR WORKSHOP: 23rd MARCH, 2007. PLEASE SUBMIT TO:  ian.parmee@uwe.ac.uk

CHAIR:     I.C. PARMEE
ACDDM LAB, BRISTOL UWE,UK.

Interactive evolutionary computing, in the main, relates to partial or complete human evaluation of the fitness of solutions generated from evolutionary search. This has been introduced where quantitative evaluation is difficult if not impossible to achieve. Examples of application include graphic arts and animation; food engineering and hazard icon design. Such applications rely upon a human-centred, subjective evaluation of the fitness of a particular design, image, taste etc as opposed to an evaluation developed from some analytic model.

Partial human interaction that complements quantitative machine-based solution evaluation is also in evidence. For instance, the user addition of new constraints in order to generate solutions that are fully satisfactory within evolutionary scheduling system or the introduction of designer-generated solutions into selected evolving generations.

Solutions can also provide information to the user which supports a better understanding of the problem domain whilst helping to identify best direction for future investigation especially when operating within poorly defined problem spaces. This supports development of the problem representation in an iterative, interactive evolutionary design and decision-making environment. Such human-centric approaches generate and succinctly present information appertaining to complex relationships between the variables, objectives and constraints that define a decision space.

In an attempt to categorise these various forms of IEC it is possible to view complete human evaluation as explicit whereas partial evaluation and interaction are less explicit, more subtle forms of human involvement. Completely implicit interaction occurs where users are unaware of their role in the evolution of a system. A simple implicit/explicit spectrum of user-centric evolutionary approaches can thus be developed.

Short papers relating to user-centric evolutionary processes across this spectrum are invited. Authors will have an opportunity during the Workshop to present the primary aspects of their research for discussion.

More details of the above with references to the examples can be found above on the Website along with details and presentations at the 2002, 2003, 2004 and 2006 Workshops. The primary aim in past years has been to provide a discussive forum around a series of short presentations by participants.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Top

 

The Website

The Website will continue to be developed to include further links to sites containing relevant information and references relating to recent papers relevant to the area.  Slides from future  Workshops will also become available.

If you wish to submit any links / references / info for inclusion on the website please contact I. C. Parmee.

 

Top

 

Links

Interactive Genetic Art

W. Latham - http://www.artworks.co.ukhttp://www.scit.wlv.ac.uk/events/latham.html

K. Sims - http://www.genarts.com/karlhttp://biota.org/ksims/

T. Unemi - http://www.intlab.soka.ac.jp/~unemi/sbart  

Fractales Web Site -      http://minimum.inria.fr/artie-fract/
                                          http://fractales.inria.fr/html/Papers/files/ps/134_lutton.ps.gz

Interactive Genetic Music 

J. H. Moore - http://www-ks.rus.uni-stuttgart.de/people/schulz /fmusic/gamusic.html

Tim Blackwell - evolutionary music . swarm intelligence - http://www.timblackwell.com

 

Problem Definition / Reformulation

Advanced Computation in Design and Decision-making (ACDDM) - http://www.ad-comtech.co.uk/ACDDM_Group.htm 

 

Interactive E-learning

http://fractales.inria.fr/html/Papers/files/ps/142_SemetHCII03.ps.gz

http://fractales.inria.fr/html/Papers/files/ps/143_SemetSIS03.ps.gz  

 

Interactive Text Retrieval

http://fractales.inria.fr/html/Papers/files/ps/133_Elise-Final-EuroGP03.ps.gz


 

Top

 

Bibliography

Bull L., Wyatt D., Parmee I. C., 2002, Towards the use of XCS in Interactive Evolutionary Design. In: Poster Proceedings of Adaptive Computing in Design and Manufacture V, Springer Verlag, London.

Cvetkovic D., Parmee I. C., Agent-based Support Within an Interactive Evolutionary Design System. Artificial Intelligence for Engineering Design, Analysis and Manufacturing Journal; Cambridge University Press, Vol.16 No.5, (2002 - in press).

Dawkins R.,  The Blind Watchmaker, Longman, 1986; Penguin Books 1988.

Graf J., Banzhaf W. (1995) Interactive Evolutionary Algorithms in Design. Procs of Artificial Neural Nets and Genetic Algorithms, Ales, France; pp 227-230

Herdy M., (1997), Evolutionary Optimisation based on Subjective Selection – evolving blends of coffee. Proceedings 5th European Congress on Intelligent Techniques and Soft Computing (EUFIT’97); pp 640-644.

Inoue T., Furuhashi T., Fujii M. et al., 1999, Development of Nurse Scheduling Support System using Interactive Evolutionary Algorithms. Proceedings IEEE International Conference on Systems, Man and Cybernetics (SMC’99); pp 533-537

Levine D., Facello M., Hallstrom P., (1997), Stalk: an Interactive System for Virtual Molecular Docking. IEEE Computer Science Engineering Magazine, 4(2), pp 55-65.

Parmee I.C., Bonham C.R., 1999, Towards the Support of Innovative Conceptual Design Through Interactive Designer/Evolutionary Computing Strategies. Artificial Intelligence for Engineering Design, Analysis and Manufacturing Journal; Cambridge University Press, 14, pp 3-16.

Parmee I.C., Cvetkovic C., Watson A.H., Bonham C.R., 2000, Multi-objective Satisfaction within an Interactive Evolutionary Design Environment. Journal of Evolutionary Computation. 8 (2), pp 197-222.

Parmee I.C., Cvetkovic C.A.H., Bonham C.R., Packham I., 2001, Introducing Prototype Interactive Evolutionary Systems for Ill-defined Design Environments. Journal of Advances in Engineering Software, 32 (6), Elsevier, pp 429 – 441.

Parmee I. C. (2001) Evolutionary and Adaptive Computing in Engineering Design. Springer Verlag, London.

Parmee I. C. (2002) Supporting Innovation and Creativity through Interactive Evolutionary Systems. Poster Proceedings Creativity and Cognition 4 Conference, University of Loughborough, CHI Conference Publications.

Parmee I. C., 2002, Improving Problem Definition through Interactive Evolutionary Computation, Journal of Artificial Intelligence in Engineering Design, Analysis and Manufacture - Special Issue: Human-computer Interaction in Engineering, 16(3), pp

Shiraki H., Saito H. (1996) "An Interactive Image Retrieval System using Genetic Algorithms." Procs of International Conference on Virtual Systems and Multimedia (VSMM’96), pp 257-262

Sims K, 1991, Artificial Evolution for Computer Graphics. Computer Graphics 25(4), Siggraph '91 Proceedings, July 1991, pp.319-328.

Sims K.,1991, Interactive Evolution of Dynamical Systems. First European Conference on Artificial Life, MIT Press

Todd S., Latham W, 1992, Evolutionary Art and Computers, Academic Press, 

Top

 

Paticipants / Interested Parties

In alphabetical order:

Blackwell, Tim;    Evolutionary music . swarm intelligence

tim.blackwell@ieee.org

Caleb-Solly, Praminda;  Senior Lecturer, University of the West of England, UK 

praminda.caleb-solly@uwe.ac.uk

Carnahan, Brian;   Department of Industrial and Systems Engineering, Auburn University, US

carnahan@eng.auburn.edu 

 

Dorris, Nathan;  Department of Industrial and Systems Engineering, Auburn University, US

Funes, Pablo;  Icosystem Corporation,  10 Fawcett St., Cambridge MA 02138 USA
pablo@icosystem.com

George, Thomas B., Technical Specialist, GE Global, Asset Protection Services
thomas.george@gegapservices.com

Herdy, Michael,   Virtuelles Prototyping
michael.herdy@inpro.de


Jacob, Christian, Assistant Prof. Dept of Comp Sc., University of Calgary
jacob@cpsc.ucalgary.ca

Parmee,  Ian C.,  University of the West of England, UK

ian.parmee@uwe.ac.uk or  iparmee@ad-comtech.co.uk

Schoenauer, Marc;  INRIA, Rocquencourt, France

marc.schoenauer@inria.fr 

Unemi, Tatsuo, Dept. of Information Systems Science, Soka University

unemi@iss.soka.ac.jp http://www.intlab.soka.ac.jp/~unemi/

Von Zuben, Fernando J.,   DCA/FEEC/Unicamp

vonzuben@dca.fee.unicamp.br

Top