Our aim in this project is to satisfy the requirements for a social decision support system, as defined by Turoff et al. [SDSS [http://www.hicss.hawaii.edu/HICSS_35/HICSSpapers/PDFdocuments/CLCSC03.pdf]Murray Turoff, Starr Roxanne Hiltz, Hee-Kyung Cho, Zheng Li, Yuanqiong Wang. «Social Decision Support Systems (SDSS)». IEEE, , Proceedings of the 35th Hawaii International Conference on System Sciences, 2002., DelphiPublic [http://www.scils.rutgers.edu/~aakhus/lap/Yetim_Turoff.pdf]Fahri Yetim Murray Turoff. «Structuring Communication Processes and Enhancing Public Discourse: The Delphi Method Revisited». New Brunswick, NJ, jun 2004. Proceedings of the 9th International Working Conference on the Language-Action Perspective on Communication Modelling (LAP 2004), Rutgers University. ]. The main aspects of such a system are:
1) A structured hypertext where participants can both explore and contribute to a set of common issues, options to address those issues, and comments to evaluate those options.
2) A dynamic voting system to evaluate those options, that can use the formal properties of (1) to identify options that respond best to the concerns of participants.
The first component is essentially an Issue-Based Information System (IBIS), one of most venerable tools for exploring consensus originating in the work of Horst Rittel in 1970 [IBIS] and now implemented in tools such as Compendium. Extensive research with the IBIS tool shows its great usefulness for eliciting consensus, and even conflict resolution [IBIS15y [http://kmi.open.ac.uk/publications/index.cfm?trnumber=kmi-05-18]Simon J. Buckingham Shum, Albert M. Selvin, Maarten Sierhuis, Jeffrey Conklin, Charles B. Haley, Bashar Nuseibeh. «Rationale Management in Software Engineering», Hypermedia Support for Argumentation-Based Rationale: 15 Years on from gIBIS and QOC, 111–132. Computer Science Editorial. Springer-Verlag, 2006., ConflictCartography [http://www.viewcraft.com/pdfs/ViewCraft_ConflictCartographyMarch04.pdf]]. However, they also warn that the formalism of the technique requires participant training, or at least a trained "secretary" to enter the data in proper form [DialogMappingIndCase [http://www.visualizingargumentation.info/]Jeff Conklin. Chap. «Dialog Mapping: Reflections on an Industrial Strength Case Study», in Paul A. Kirschner, Simon J. Buckingham Shum, Chad S. Carr, Eds. Visualizing Argumentation. Springer-Verlag, London, 2003, pp. 117–136. ]. Neither of these conditions scales well to a population.
A more fundamental issue with this formal design, beyond training, is that the very definition of common issues and purposes may be what is at stake, and in that case the more formal mapping out of issues might not tell the whole story. Indeed, the most successful stories of IBIS implementation have incorporated that formal approach into conventional face-to-face meetings; and those meetings always involved an informal storytelling element, which was then formalized. Although I fully agree with IBIS practitioners that the formal representation of issues, options and arguments helps neutralize conflicts around those stories, I also believe, however, that this formal representation could only fully accomplish this after the story underlying it has been heard. Also note that IBIS has been used more often (though not exclusively) in corporate context, where a common purpose can more easily be found than in society at large.
Both those considerations persuade me that, though a formal IBIS representation of issues is a desirable goal, capturing free-form contributions is essential to an inclusive, large-scale consensus-building tool. But free-form contributions, as they already exist in forums and blogs, do not allow consensus on their own: The tool must also assist the subsequent transformation of free-form contributions into structured contributions. In my opinion, such abstraction cannot yet be reliably automated, and I suggest it must still be done by trained humans (though text clustering algorithms could possibly make relevant suggestions, tentatively associating a given contribution to one of the formal elements already identified in the system.) Hence, I would advise relying primarily on some trained participants collectively structuring new proposals as they come. (This is not unlike some descriptions of social annotations for the semantic web, as in [CollTag [http://www.inf.unisi.ch/phd/mesnage/resources/pdf/mesnageJazayeri06b.pdf]Specifying the Collaborative Tagging System, Cédric Mesnage and Mehdi Jazayeri, SAAW (2nd Semantic Annotation and Authoring Workshop) co-located with ISWC, November 2006.]) Doing this necessitates the ability to mark runs of text in other participants' informal contributions, and to associate them to the formal hypertext structure; and the ability to review and validate those associations, possibly with the help of a moderation system. It will be possible to identify quickly those contributions whose contents have not yet been integrated in the formal system, and conversely participants whose contributions were woven into the formal structure will know they have been heard.
The incentive for trained participants to undertake this demanding structuration task, beyond the possibility of discovering new options, is to find contributions that they can validly associate to their own favoured options, thus adding to the votes to that option, subject to review and approval of the association by the original author of the contribution. Also, this may encourage trust links between expert and novice users that could either lead to mentoring opportunities, or form the basis of a liquid democracy system, as proposed in [LiquidDemocracy [http://fc.antioch.edu/~james_green-armytage/vm/proxy.htm]].
The intent is also that, at all times, the formal definitions allow to refer back to the stories contained in the informal contributions. Just as a voting system may help to resolve options in the formal structure, we expect groups to form around certain options, and to review the stories around them so as to present them in the most forceful manner. Thus, the narrative and informal contributions should converge in parallel with the formal, and hopefully this may help achieve better understanding of all participant's diverse viewpoints, beyond simply reaching a common course of action, however needed.