Generating Tag Fields for Individual Patterns: Networking our Patterns

After my evening walk of contemplation I began considering more deeply the idea of tags and providing more powerful ways to express interconnected meta-data as related to the Liberating Voices patterns.

Currently, the L.V. system uses a sort of quasi-tagging approach that defines categories and topics. For the most part these “tags” serve their proposed purpose. However, from my perspective it seems that the lack of dynamic capabilities minimizes the potential for filtering and organizing patterns. Instead, perhaps we should be thinking about ways to enable dynamic tagging where users can generate a folksonomy surrounding patterns.

This doesn’t mean we should get rid of what is already there, but rather we should extend the functionality of the system. It seems that the current terms being used are very high-level and generic. By keeping these more generic terms we can perhaps help frame the types of more user specific tags.

However, rather than merely providing a pathway to develop a folksonomy for the patterns, we must remember patterns exist in an often different way than a lot of content. For instance, patterns are keys or building blocks that exist with a multi-dimensional structure. In this way patterns possess relationships along a vertical as well as horizontal axis.

With this in mind perhaps we would instead want to start defining the levels of abstraction that patterns have by numerically describing the various levels. In this way a cluster of patterns would exist at a certain level. Now, L.V. has already categorized or clustered patterns but I do not think this is an explicit field in the database model. I think we need to add it!

To make this usable from a users standpoint, maybe a radio button describing the levels/categories and then some ajax call to the database would pull up all the tags related to a certain level, perhaps you could then define your own tag for the level within the hierarchy. Maybe we want to be able to annotate our tags and discuss the reasoning behind the categorization of the data in such a way. This annotation could in a sense point to related cases where the pattern was used and likewise show the relevance of the tag.

Why is this important to me? Well, because I am in the process of developing a listing of patterns related to my work, but that correlate directly to a previous post. I'm right now looking at education patterns as related to increasing information literacy and knowledge for natural resource management. An alternative and more tag friendly environment as well as way to search and filter through these tags would be great.

Fortunately, we have something like this in its infancy, but I think it will become a priority of mine in the near future as I attempt to build systems and processes to construct pattern languages based upon the L.V. repository.

Wow, I’m not sure if that made any sense, so I think it is worth a use case/interaction diagram at least.