Showing posts with label Influence Diagrams. Show all posts
Showing posts with label Influence Diagrams. Show all posts

Friday, April 4, 2008

Personal Brainstorming Session

About two nights ago I got a wild hair and walked to the office late at night and began going at it on the white-board. During this process I clarified several things.

1.) Defined an assumption about ICT's need to reflect community praxis, rather than ICT redefining praxis in haphazard ways.

2.) Articulated the components that make up the theory and practice of the work I'm focusing on.

3.) Wrote out a preliminary model for the architecture of the system I'm working on to support this work.

The images below reflect the evening's work.








This summer, is going to be devoted to building the base application, along with coordinating with potential case-study participants. Busy times...

Sunday, March 30, 2008

Need to build some software...

While I prefer the use of influence and causal loop diagrams to the approaches of mind-maps and concept maps, the issue with the former is that there is little in the way of free and opensource tools available for using this approach, or learning the process.

There are tons of applications devoted to influence diagrams, system dynamics and so forth, but they cost hundreds, if not thousands of dollars. There is one opensource modeling program out there but it does not include influence diagrams.

This might be primary reason pattern adopters have been slow to consider such an option, because for many it isn't an option; it certainly isn't one to me. So I guess the option I do have is of course to code my own application.

I don't really have any intention of making it full-featured. I just want to be able to add system elements, embed text (to contextualize the element, with simple descriptions shown onmouseover()), use arch lines to show relationships, add (+) or (-) to differentiate the type of influence, define the system boundary (maybe define a meta-system), be able to add images as user-friendly icons for elements, and I want to be able to have versioning (that way the model can be adapted to reveal changes within the system that might warrant additional responses).

For now, I'm going to build a crude prototype, meaning no AJAX. I don't have the time to figure it out at this point. As in a past post, I still plan on using SVG because of it's dynamic options compared to the php_gd2 option. Plus, I have taken some serious time figuring out the SVG option and I like it. It gives me an option to add AJAX capabilities later on. Ultimately, I would like to code this for my own research and if someone finds use, then let them add the bells and whistles later on.

Anyway, I think I will draw up a more complete list of requirements and then fire up my agroUML app and get to mapping this project out.

Note: I really hate not having time to fully engage cool projects like this. Yet, I have no intention of being a full-time web developer again. Well, unless it is working on building apps to support my research interests. Hmmmm, potential dream job?

Saturday, March 29, 2008

The process of visualizing complex environments

In order to make patterns an effective approach to any community empowerment project it is necessary that planners, community members and other stakeholders understand the social, political, economic and environmental context in which they are working within. This perhaps the case in any situation regardless of whether patterns are the model you use or not. Questions must obviously arise to help understand power relations within the system, as well as the nature of information and communicative influences among actors within the system. Likewise it is necessary to evaluate the influence of resources, such as money, access to technology, levels of social capital.

The list below shows areas where project planners, social analysts and community members might want to consider in the construction of a visual model that represents the current context and forces associated within a community system. Though it is not exhaustive it does represent some important considerations.


Trust
Cooperation
Natural and Human Resources
Information – Production and Consumption
Health of Environment (Ecology)
Economic Capacity


Education – Literacy
Political Influence
Social Capital – Affiliations
Health
Technology


Along with this list it is also perhaps necessary to define the nature of these influences by structuring this model to include different levels of analysis.
  • Global-Level Forces, e.g. influence of FDI, multinational corporations (MNCs), International Norms and Treaties, influence from adversarial or cooperative nation-states, transnational advocacy networks (TANs)


  • State-Level Forces, e.g. nature of governance (political opportunity structure), presence of corruption, oppositional policies, media propaganda, elites,


  • Community Level Forces, e.g. community divisions, resources, education, cooperation vs. fragmentation, gender relations, effectiveness of leaders (see list above)

Once a list has been generated among participants, the next task is to begin understanding how each of these elements link to each other to construct the context. There are several ways in which this can be done. For instance, Joanne Tippett from the University of Manchester uses a modified version of mind-mapping to elicit insights into the contextual nature of a community system (yet with an emphasis on design). Others use concept mapping or interaction diagrams borrowed from the UML, popular among software developers.

However, it is felt that simple influence diagrams can be much more revealing in showing multiple interactions across multiple dimensions. One can define the elements in the system, how they influence one another, as well as define the boundaries of the system. Users can also identify elements from outside the defined system, which can be useful in understanding the potential effects of second or third-order influences on a system. An example of one these influences might be the role of the Universal Declaration on Human Rights, for framing grievances against state abuses of a particular population.

(Source: UK Open Systems Society)
Yet, despite the potentials, influence diagrams lack the visual appeal of mind-maps. Though it is unclear whether this is a problem, it is possible that developers could extend the influence diagram model to include pictures or other images in order to pull these diagrams out of the antiseptic feel they currently possess. Nevertheless, these models do provide a tidy methodology and visualization approach that can be useful in understanding complex social environments and thereby perhaps useful in the construction of pattern languages as responses to these environments.

While this has yet to be tested with patterns as central to the overall process, the use of influence diagrams have been relied upon heavily in the application of qualitative system dynamics and therefore have a history of success. The question then becomes, can we use influence diagrams in configuring pattern languages and can these models be useful in non-expert contexts where multiple stakeholders with varying levels of literacy can in fact make sense and use of these tools.

This of course requires further investigation