Musings & Meanderings

Has this already been done or tried?

Can a traditional hierarchical structure be automatically modified by searches and, separately, by tags? If there were set thresholds inherent in the modification instruction, such as 10 or 100 either search word combinations or tags attached to the same image or word combinations, would that address the random clutter problem? Could potential risk to the database(s) be addressed sufficiently through security filters?

The related question is how many and how wide a range of different modification rules and rights could be assigned for a single data structure. If you have a dozen or more classifications of users with different access and privileges, does each of their input enter via a different track, or can segments of it be pooled after leaving the space where the user status is defined and protected?

Credentialed users, for example, could create sub categories without limit, make multiple and faceted entry of items, etc. New users, at the other end of the spectrum, could make their own tags and links in their accounts and these would form a component of communal classification which would be automatically collated via being pooled. Between the two, various levels of intermediate and non-tech expert users could be given appropriate levels of access and rights, and so on. So I’ve imagined it, anyway.

Giving a non-tech user a way to make faceted entries would obviously have to be done in common language rather than jargon. As an example, apples can relate to (beyond food and fruit) cooking and dessert and even biblical symbolism. Sorting things from one’s own special interest list could be both easy and popular fun. There’s an assumption in this concept that a user putting a ‘gemstone’ tag beside an apple picture or article is going to be an aberration. This example does, however, indicate a requirement for defining category levels and relationships between them, as apple could easily be found at a different level down a sculpture or jewelry branch.

Many, perhaps even all, of the different components in my imaginings already exist, but the applications of them, in my experience, are usually very limited. It sometimes feels as though every basic body movement (in analogy) is already enabled, on a computer, over the internet, on a website or through a web app. Continuing the analogy, I have to go over there and sign in to lift my little finger, and somewhere else to walk, and somewhere else again to sit down. Spending too much time on the web is beginning to make ‘me’ feel like a jigsaw puzzle that no one has assembled, so that fragmentation of my identity results. Enterprise applications, on the other hand, make me seem (to myself) rigidly 2-dimensional, like a cardboard cutout figure. Sometimes, if they work well, the image is of a paper chain of figures. Not exactly fragmented, in this case, but rather constrained within a 3rd party’s narrow definition.

Although I specifically made the original question in this post about a single traditional data structure, I am also naturally wondering about applying it to relational databases together with ontological meta-tagging. The first focus of my thinking, though, remains on how to begin achieving balance between authoritatively compiled data and user generated data, while retaining the maximum value of both.

This entry was posted on Tuesday, December 5th, 2006 at 10:28 pm and is filed under Web Development, Semantic Web. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.