Tuesday, December 09, 2003


Have just started a seminar on Excellence here in London. Excellence is a concept of interest to this blog as an understanding is key to enabling the discovery of excellent resources, such as songs by unknown artists. Current thinking on this topic is outlined below, although I would expect to gain a greater degree of clarity as the seminar progresses (so far we have had session one of ten). There appear to be two ways of incorporating the concept of excellence into Internet resource discovery networks as far as I can see.

Firstly, there is the rating of songs. To do this an implicit scale in which the top category can be labelled "Excellent" is used, although other superlatives such as "Outstanding" or "Very Good" are also viable. This is effectively the conceptual framework behind any rating process (where some resources are ranked higher than others). The rating of music in this way would not, however, appear to be an inherently satisfying activity for users as it is in essence a ranking or judging process - a view that was touched upon in the seminar.

So, for example, the posting of a "Good" rating for Open Letter by Horton's Choice (see last post below) was not all that rewarding for the parties involved. Potential users of a resource discovery network are not looking for songs with a rating of less than excellent, and as a rater I would prefer for all songs on the network to be excellent, however difficult this may be to achieve. I have not even mentioned how artists must feel when their song is rated below the top category. Indeed, I would expect artists to be somewhat reluctant to even enter such a rating process for the first time (although they may be pleasantly surprised if they were to achieve the top rating). The whole range of ratings (Excellent, Good, Neutral, Poor and Terrible) are, however, vital in enabling collaborative filtering to occur (I will post an analysis of collaborative filtering and explain its part in the discovery process next year).

A second way to consider the concept of excellence is to examine the relationship between resource and rater, and see that something is only "excellent" as a result of being declared so. This is because nothing is intrinsically excellent; it is only excellent in a persons' opinion and as a result of having been declared so by that person. That is why the same resource (song) can be declared excellent by one person and poor, or even terrible, by another. It also follows that nothing is intrinsically "Terrible", it is only so in relation to having been declared so by somebody.

This second view is potentially far more powerful for participants in a discovery network (through enabling people to declare that resources are excellent). Not only can it enable excellent resources to be incorporated into the network, but it could also be a far more rewarding activity for those involved, when compared to the first rating process outlined above. By requiring that songs are first declared to be excellent by a user, in order for them to be included in the network, it is more likely the network will achieve the goal of being populated by excellent songs.

The possibility that artists will have their music declared to be excellent is also a potentially rewarding experience for them, and the user declaring it to be excellent will be aware that they are adding something of value to the music, enabling others to find it.

Overall, the possibility of enabling the discovery music through the declaration of excellence, on a global scale, seems to be a far more appealing implementation of the concept of excellence, when compared to that of simply rating music. This is what I am committed to creating out of being in the seminar.

More to follow...

Comments: Post a Comment

This page is powered by Blogger. Isn't yours?