The following day, after many wasted pleas to meet my tee time, we headed to Wal-Mart to purchase the storage crates. They had green ones, blue ones and clear ones; ones with wheels and those without. All shapes and sizes were available, but we settled on a couple ones that seemed to fit the type of materials we needed to store. On the way home, we stopped at Staples to pick up a labeling machine which would allow us to place specific classifications on each box, crate, folder and cabinet.
Back home, the process of cataloging began by noticing that we had three initial high-level classifications: Christmas, travel and computer parts. As we moved to organize the Christmas objects, three additional sub-classes emerged: tree, china and decorative. Three crates soon became four when the ornaments wouldn't fit into a single crate. Interesting, the tree ornaments were eventually super sub-classed into glass, traditional and kids. Not sure where the yarn angels go but I slide one to the dog and proceeded to create a new class called "throw away." This process continued with travel objects (sorted by geography), and one would think that the computer equipment could easily be divided into various sub-classes. Unfortunately, the house manager had to attend to the two year old, and I was left in charge of the home taxonomy. No problem here, cram everything into a single box and what ever doesn't fit is thrown away. The reason why I was "impeached" as the house manager many years ago. Upon returning, we continued with the rest of the objects and created, grouped, divided and collapsed the different classes. Everything is in its place, easy to find and most of the real junk tossed out. What does all of this mean to us in the meta data world? Basically, the events provide us with some insight into how to build taxonomies, at least on a small scale. The following paragraphs describe some of the lessons learned.
Without a doubt, the most important decision was to bring in an expert that really cared about the results. The house manager was the person for this job; her personality is a true "A" type that organizes everything. She ensured that we stayed focused and that the end result was usable by any member of the team. She didn't get lost in building something enormous in size with custom made cabinets. The shelves were simple and easy to build; we focused on delivering the value that the project demanded. The skills required included foresight to see what assets we had and the very best way to organize them within the environment we were working in.
Figure 1: Poor Drawing of Home Storage Taxonomy
Realize that organization and taxonomies take time to build. Two weekends were set aside in order to complete the application build (shelves), test (check the weight capability and size) and data load. Could the shelves have been built quicker? Of course, but the end product would not have been as usable and organized. The true value was the evolution of the taxonomy as described in the earlier sections. Could we have gone to "Shelves-Are-Us" and purchased preprinted labels and containers? Yes, but the majority of labels would have been tossed out, representing a waste of time and money. More importantly, would that taxonomy be as useful to the end user as the one we developed? The taxonomy was not producer based but rather consumer based. Consumer-based taxonomies are emerging in the technical world and are better known as Folksonomies. A Folksonomy is a flat namespace where there is no hierarchy and no specific parent-child relationships [Mathes, 2004]. Users can classify their content with simple keywords such as customer, order, schema, service, building, etc. This kind of classification works great when the user has a limited set of assets they are interested in. Depending on your job description, organizing technical assets in a Folksonomy works great since the number of assets is usually fairly small. For example, the organization may create a taxonomy structure for a customer name service as follows:
For a developer within the enterprise services organization that works within the customer profile segment, they may file the service under "Name." Within their context or job, this classification makes perfect sense. Currently, users of the repository already have a simple classification system outside of the repository in the favorite's utility inside the browser. Unfortunately, that information cannot be accessed by the repository group in order to build additional classification methods. Technorati is an example of an online service utilizing folksonomy technology. They provide a search engine and classification taxonomy across the collection of blogs. A Weblog or blog is personal journal. Weblogs express as many different subjects and opinions as there are people writing them. Some blogs are highly influential and have enormous readership, while others are primarily intended for a close circle of family and friends. Technorati provides a meta data search engine based on the RSS feeds which is the method of publication for most blogs. The reason I bring this up in the context of this article is how this application has altered my classification or taxonomy system for my own blog. Recently, I went back and reclassified all 100 entries based on the system used within Technorati. For example, I was classifying my thoughts on the future of technology under "The Future of Technology" which sounds reasonable. However, everyone else was using globalization, future, technology or innovation.
Be the first to comment on this post using the section below.