This book is primarily a research monograph, in which the discussion
of the main topics has been broadened so that they are related to their
surrounding context in information retrieval as a whole; it is not a
textbook, and no attempt has therefore been made to justify the choice
of topic, or account for the use of certain concepts, or to provide an
elementary description of either. For instance in Chapter 1, it is assumed
that the reader is familiar with the idea of using keywords in information
retrieval: I have not considered the relation between this kind of
retrieval device and a controlled thesaurus or descriptor set, or that
between the use of simple class lists as document descriptions and the
use of descriptions with a syntactic structure, for example. Equally,
in Chapter 2, I have made use of recall/precision ratios as a means of
characterising retrieval performance, without justification or argument;
but this does not mean that I am unaware of the difficulties of doing this,
or of the attention which has been devoted to, and controversy which has
raged round, this subject; it is simply that from the point of view of
my main purpose it is reasonable to use these ratios.