All this is not to suggest that traditional KOS
have become irrelevant in the present-day context. It
is only to suggest that they have limitations in terms
of the space/information environments within which
they can operate and be effectively employed; we
need to look for more scalable KOS to complement
traditional KOS for the digital environment. This is
not something entirely new and the need has been
recognised for quite sometime. let us re-state our
problem and requirements and illustrate these by
means of a couple of examples. The amount of
data available online has grown and we need to be
able to search large-sized collections in the order
of billions to trillions of words. There is a need of
more flexible matching (with query) facilities; e.g.,
to be able to search for texts in which the word
‘INFORMATION’ is near the word ‘RETRIEVAL’ and
to be able to define ‘nearness’ according to our
requirements. In view of the volume of data, it is
important to have a ranked output (Decreasing degree
of relevance to the query). In other words, a KOS
that is capable of informing which of the retrieved
documents are more relevant. Some examples for
the same are: