About categories, tags, keywords, metadata and so on...

Semantic descriptions in the library

Which is the difference between categories/tags/keywords/metadata?

There are of course small differences, for example:

metadata can be read by machines
categories tend to become standards
keywords can be referred to particular words in the text
metadata can be read by machines, and when edited the change is embedded in itself
tags are metadata without history
and so on...

Despite those technical differences, in the context of the library, they can be grouped as semantic descriptions of a book (material book + content).
Within this framework, the semantic description allows organizing the library following a certain pattern, but...

Phisical & Digital / Universal & Particular

In the physical library...
is possible only one organizing pattern at a time.

In the digital library...
multiple organizing patterns are allowed, you can just select which one do you want.

Furthermore, in the world of semantic descriptions, there is an ongoing debate on the universality of such organizing patterns (tangled within an ideology of absolute truth), VS subjective descriptions.
Physical libraries and librarians seem to embrace this kind of universal ideology. Due to their impossibility of having multiple organizing principles, they would like to organize their shelves as best as western idealism allows. Here standards born, as mushrooms...and they fight to become standards...and all you can say about agonism and hegemony, democracy and dictatorship, subjectification, control, and freedom.

In the digital library, as well, the translation of this debate follows the same paths...but on the other hand, the web itself allows not only to have multiple choices, and so multiple organizing principles at a time, but also to increase the level of participation and democracy.

A social network to challenge semantic descriptions

How can the web merge the interest in universal categories without to lose the need for a self-driven subjectification?
The response to this question can be found in the behavior of contemporary social networks (in the broad sense). In those public spaces, it is becoming a praxis to allow a review of contents, as well as ranking them.

I would like to dream about a system that allows using rankings and reviews to build a sane democratic agonism purged by any absolute hegemony and dictatorship. The semantic description should be intended in its double value: the one who tends toward a universal standard, and a second which allows the subjective perspective.
It is essential to highlight how the hierarchization of such semantic descriptions is not intended to hide the one on the bottom of the pyramid, but just to show (if required) the one upon which there are an agreement and an open debate.
To allow this kind of system we must follow two principles: openness and temporality.

Openness/Inclusivity, allow having all kinds of subjective description, as well as jokes/trolls/...
Temporality, allow formulating new agreements over time, escaping from stagnation and control...

Openess/Inclusivity

The concept of openness is necessary because everyone can personalize its organizational pattern through personal semantic descriptions, as well as to find new organizational patterns through the lens of others eyes, other personal descriptions.
I tag different texts with the word 'papadupa'. It would be hard pressed to even find someone agreeing on my tag, so in the chart of ranking, by any chance, it would be on the bottom level. But nevertheless it would be visible and someone interested by the world could visualize which are the contents tagged with this word augmenting the level of serendipity, which otherwise would be a boring standard.

Temporality

The temporality aspect is meant to be a recurrent moment in which the agreement is re-evaluated. It allows us to don't rely on a single moment in time, where the evaluation first started, but to have multiple moments of evaluation (for example one at months or a year). This kind of retreatment of the ranking principle can be helpful to highlight and compare the different moments in time, as well as be used to constitute a dataset and define the general behavior of this kind of system. Nonetheless, it allows keeping the debate on single keywords alive and prove their effectiveness in time.
I tag a text with the word 'xxxxx'. The word starts to be ranked positively and goes in a top position. In the meanwhile, a debate starts to address the problems of tagging that text with such a word. Here the word can move up, keep its position, or move down, in the chart based on the ranks.
After a certain period, a new session where the words can be ranked from zero starts again. That word can be ranked again and move to another position than before. The debate can continue.
The past session of ranking tags is still available and you can compare it with the new one. After years, you will have many sessions to compare and define the fluctuations of that particular tag.

Others principles

multi-mediality > everything can be considered as a tag to rank and debate
recursivity > tagging tags

User:Tancre/Special Issue 9/About categories, tags, keywords, metadata and so on...

Contents