Lumping stuff together ;-)
One of the things we appear to do, as
human beings, is classify things. We use this both to tag (label)
things so that we can work with abstract models, and as a way of distinguishing
between things.
Lumping together, and separating.
Using the same basic technique.
We assign labels to things, because
fundamentally we are pattern recognition engines. We see, smell,
touch, hear or taste something and it reminds us of other things we have
sensed. If we read a description of something, it builds a model in
our minds, and we associate that with other models we have built before.
Most aggregators, of course, don't really
even aggregate. They present a number of different things in one
tool.
|
Classification
It's all about similarity
and difference. But it is also about finding a hyperplane in
feature space, or a function of feature space, which classifies your
target objects in the problem space. And it might be about doing
that optimally, in some sense - either the maximally separating
hyperplane, or optimal in terms of time taken.
Where you have human agents
classifying for you, such as with social bookmarking, the problem
changes. Now we have some idea of the classifications to use, but
the nature of communal decisions is such that we cannot be sure of the
'correctness' of the classifications used.
But who gets to define
'correctness'? There is a dislike, on the part of philosophers and
logicians, of the act of appealing to the populous for their
opinion. This can be viewed as stratification based on proven
ability, or, depending on one's point of view, academic snobbery.
Folk psychology and similar, is therefore often considered to be inappropriate
in an academic setting.
Folksonomies, however,
being the taxonomies generated by the 'folk' engaged in social
bookmarking and similar activities, are the subject of a reasonable
amount of academic study.
|
Trends
One thing aggregation lets
us do more readily is absorb the volume of information which is being
produced in the world today. For instance, there is a huge amount
of information being produced about social networking. There are
tools around which can help us reduce these to more a mangeable amount of
information, from aggregators which summarise blogs on a particular
topic, through news services and alerting systems, to services which
present us with summary information on the amount of web pages and the
like mentioning a subject (see, for instance, Google
trends)
|
