Background Study
Project status: The project ICC was initiated by the University of Luxembourg. It started April 1, 2006. The City of Luxembourg acts as a co-partner (test bed).
The city plan is to create a
digital city by transforming their current city web site into a centralized online administration the
e-City integrated into eLuxembourg. The goal is to provide citizens and visitors with a suite of electronic services as well as to give them online access to the main official forms generated by the city and interact with them. This is part of the eEurope strategy as first defined in Lisbon (2000).
Conviviality: There is neither a clear model nor any singular vision of conviviality and the definitions are domain specific. Generally speaking, a convivial place or group is one in which individuals are welcome and feel at ease, for example a city portal with text in the languages spoken by the constitutive communities of the city. There are therefore many kinds of conviviality.
In
[1], conviviality is a social relationship based on the acceptance of others. It describes a multi-agent recommendation system that finds matches between students in difficulty and students who can help them as tutors. It uses a utility function based on “structure measures” such as students’ social profile, sociability degree, mood states, acceptance degree.
In
[2], reputation is found to be a necessary condition for conviviality. This work defines a functional ontology of reputation for multi-agent systems. It defines reputation roles played by agents to evaluate, propagate and receive reputation.
In
[3], Conviviality is not the summ of local characteristics (criteria) but a global chracteristic that emerges from the system. A convivial agent is rational and cooperative. This means it can negotiate with users, offer responses and suggestions based on contextual interpretation, is flexible when interacting with users and allows their contestations and feedbacks, it offers cooperative reactions to users’requests anticipating their needs and finally offers responses in a form useful to them.
Conviviality as a set of constraints on behavior following social conventions can be considered as a norm.
Research questions
How to model the social concept of conviviality? What aspects will we address? How to apply such model to digital cities?
Scope
Digital cities are multi-facetted and complex entities. Our focus will be on the city web portal. The obvious question is to evaluate the model experimentally.
Sub questions
- What is the role of conviviality in literature?
- What is the most appropriate model for conviviality?
- How to define measures?
- How to evaluate the model?
Motivation for the city
Users go to city web portals to fulfill needs they have (information gathering, filling up a form, entertainment, etc.). However, while trying to reach their goals, users have to fight against constraints and must find compromises.
Constraints can be:
- Imposed by the needs or behaviors of other users (system use, misuse, abuse, etc.)
- Imposed by the objectives of the city administration (control over information, identification requirements, etc.)
- Technical (system overload, lack of functionality, etc.)
Application in CS/agent context
Agents assist users in getting their needs satisfied in the best possible ways while taking constraints into account.
Methodology
- Write a literature survey of conviviality
- Create a multi agent organizational model of conviviality (AGR [4], Moise+ [5], etc.)
- Define criteria for specific contexts (administration, themes, etc.)
- Apply test cases from the city of Luxembourg to the model
Examples of criteria: Security and trust; Interface flexibility and adaptability (personalization, customization, etc.); Ergonomics (web site design and heuristics); System scalability, efficiency and speed; Density of users in any given location, groups stability.
Benchmarking data is expected from the run of use case scenarios experiments on the City of Luxembourg portal.
Success criteria
- Generality of our model
- Evaluation and measures of conviviality
Expected problems
-
How to avoid reducing conviviality to one of its components? How to preserve its core value and meaning? For example, a problem with formalization and implementation is that the concept itself is inherently non-formal. If you start formalizing it, you end up with rules or norms or trust or reputation, brownie points, or other such cold economic notions that miss the point. It is quite likely that on analysis, conviviality disappears and is reduced to other notions.
-
What is the common ground between users? Who are the users? How to gather information about users and establish some kind of profiles to define conviviality criteria and adapt it to users and groups?
-
How to define measures such as individuals’ mutual acceptance?
Possible collaborations
With for example a social sciences research group involved with similar empirical research to address questions such as:
-
How do ICT tools support conviviality in communities of practice? Groupware tools evaluations and research (ex: Erickson and Kellog concept of “translucency”)
-
How does conviviality affect coalition formation among humans? Motivates group formation and development; Motivates learning; Reinforces social cohesion
-
How does conviviality affect knowledge sharing? Encourages cooperative behaviors; Reduces mis-coordinations that result from breakdowns in shared knowledge
References
PhD. Work plan: 3-year Timetable
First Year: March 2006 - February 2007
Research
- Domain definition
- State-of-the-Art and background research:
digital cities;
Conviviality
- Formalism
Implementation
- A conviviality scenario using agent systems (AGR, Agent-Group-Role/Madkit; Moise; …)
Simulations
- Testing conviviality scenario with agent systems (AGR, Agent-Group-Role/Madkit; Moise; …)
Publications
Accepted papers:
-
WBC 2007: Full paper. Web Based Communities. Salamanca, Spain, Feb. 18-20
-
ASAmI'07: Technical paper. Symposium on "Artificial Societies for Ambient Intelligence". AISB Convention. Newcastle, April 2-5Submitted papers:
Upcoming paper submissions:
-
SID 2007: 6th Intl. Workshop on Social Intelligence Design. Trento, July 2-4 (deadline: March 1)
-
Ludovia2007: International conferencence on "La Convivialite des interfaces a vocation ludique et/ou pedagogique: Conception, creation, valeurs, usages."Ax les Termes, Arrieges, July 4-6 (deadline: March 1)
-
CSCW'07: ACM Conference on Computer Supported Cooperative Work (info TBA)Additional possibilities:
-
GROUP '07: ACM 2007 International Conference on Supporting Group Work. Florida, NOV. 4-7 (deadline: TBA)
-
SAB'07: From Animals to Animats 9. The Ninth International Conference on the Simulation of Adaptive Behavior (info TBA)
Colloques, confererences
-
WBC 2007: Web Based Communities. Salamanca, Spain, Feb. 18-20, 2007
-
Dagstuhl Seminar: Dagsthul, Germany, March 18-23, 2007----
Second Year: March 2007 - February 2008
Research
- Environment, context, structure
- Group: Organization, equilibrium, set hypothesis for the added-value of conviviality
- Individuals: social distance, goals of individual vs group, roles
- variables and measurements to compute conviviality
- Formalism
Implementation
- Conviviality scenarios using agent systems (AGR, Agent-Group-Role/Madkit; Moise; …)
- Running Prototype
- Model interaction with applications
Simulations
Testing using agent systems (AGR, Agent-Group-Role/Madkit; Moise; …) with ICC scenarios.
- Conviviality rules: definition, enforcement, relaxation
- Group organizations: verify hypothesis
Publications, Visits
- At least one paper submission
- Workshop Participations to accepted 2006 submission
- Visits
Third Year: March 2008 - February 2009
Research
- Finish Research: conviviality model ICC
- User induced conviviality
Implementation
- Finish agent systems implementation
- Experiment with eventual additional research findings: user induced conviviality
Simulations
Test scenarios using Luxembourg city as test bed
Thesis
- Write and finalize thesis document
- Defense presentation
- Publication TBD
Publication
Article summarizing ICC model and prototype
Short Curriculum Vitae
-
03/2006 - today: PhD Candidate and teaching assistant, Luxembourg University. Special Interest Topics: "Inventing Communities of Communication", Conviviality and Digital Cities.
-
1999-1994: Master of Science in Computer Science under Dr.jack Schwartz, New York University. Research assistant at NYU Media Lab. Special Interest Topic: "Non-verbal Communication over the Internet".
-
Previously: Master of Fine Arts, Angewandte Kunst Akademie, Vienna. Bachelor of Fine Arts, ENSAAMA, Paris. 6 years industrial experience (Stanford Research Institute, Scient, Netscape and AOL).
Member of IEEE, ACM, SIGCHI
This work plan fits my particular expertise as it builds on ... (ex: organizational model, java, etc.)
Books & References