From Metadata to Folksonomy, a new way of thinking about content.

What is a Metadata?
Let me first refresh your memory about a little bit of definition. A so called “metadata” is data about data. Not very clear isn’t it? It is actually a set of data or information about another data, for instance a file or a set of files. Metadata are everywhere, on internet, in your MP3 player, in your holiday photos and videos… To illustrate that in a more concrete way, we could take the example of an MP3 file on your Ipod. You have probably noticed that in addition to the music, you can also experience some value added information like the name of the artist, the name of the song, the type of music and even sometimes further data like a rating or the cover of the album… These are all metadata.

One of the more interesting issues about the metadata is the fact they are stored in databases. The matching data can thus be searched through all metadata information. Taking the above example, we could very easily look for all the pictures which have been taken by a Canon EOS 450D camera or/and on 12/12/2009… every database fields are searchable! We can easily guess the endless possibilities such data can provide. As these tables are only filled with texts or numbers, compatibility issues are lowered and it helps a lot especially when speaking about cross media or cross platforms usage. Metadata are mainly used by software and other search engines to identify contents and then exploit them by processing the content.

SOS, Data are surrounding us!
We can compare the growth of the average number of files on a computer to the increase in power of microprocessors. A single modern computer can now manage the same amount of data as an enterprise server 10 years ago! Both digitalization and globalization have subsequently increased the amount of files managed by a single person! Just try to figure out how many photos, videos or songs plus all other work related documents you are dealing with every single day?
When listening to Computing and ICT professionals, the explosion of the amount of data is THE MAJOR ISSUE. The biggest challenge is now to find efficient solutions to classify and make all these information searchable. Folksonomy or Taxonomy as a part of these metadata could be the future of sorting.

Taxonomy and Folksonomy :
Ed: Please do not flee when hearing these two latter words, it is not that difficult, a proof is the following video:

Click to play the movie

Not so difficult, isn't it ?

The term taxonomy dates back to Darwin’s classification of species. It is following the same logic of sorting. For example in a category named “Movies” we can find other subcategories like “Comedy”, “Drama”, “Action” and so on. This principle is used in computing and in databases management as a way of sorting data.
Folksonomy is an evolution of the Taxonomy without any hierarchic relations categories or words. It is a key word approach which allows people to freely label or “Tag” contents. The users will be able to search contents by using key words. This definition heads us to another major issue about sorting and content classification: matching the Folksonomy with content.

Matching Folksonomy with contents:
The Folksonomy has to describe the content with the highest level of accuracy and reliability possible. The major problems we can encounter are on the one hand the human factor and on the other hand the necessity of human in the matching process:

Human factor: Content can be described differently by different people depending on their own experiences.

Dependence on the human factor: let’s take the example of a picture from a painting, as artificial intelligence does not exist, a person will have to label the picture with all the uncertainty explained in first point.

Hopefully Internet allows people to communicate and act in a collective way. As in statistics, the more the size of a sample is important, the less error you will have. The same logic can be fitted to Folksonomy, the more people are tagging content, the more accurate will be the matching. The collective intelligence notion is thus a fundamental part of this new way of thinking about contents.

Before leaving...
The field of application of Taxonomy, Folksonomy and generally speaking Metadata is wide. It could be used in a personal usage like when tagging a friend on a photo on Facebook or when rating your favorite song in ITunes. In enterprises, tagging is used to sort information in ERP, CRM or intranet systems. Metadata could embed for instance digital signature or versioning information…

Ed: I hope after reading this post you will now reconsider using your long winter evenings to clean your multimedia libraries instead of playing Scrabble! Good Luck...

Author : Frédéric Kretzer