Target word clustering takes as input multiple contexts, each of which
includes a single target word that is marked with a special XML tag
called "head". The object is to cluster those contexts to discover the
different meanings of the target word. When using LSA, the premise is
that contexts that contain words that have occurred in similar contexts
should be clustered together.
When using LSA to carry out target word context clustering, each feature will be represented as a vector that shows the contexts in which it occurs. Each context will be represented as the average of the vectors associated with the features that occur around the target word in that context. The premise is that contexts that are made up of features that occur in some of the same contexts should be similar to each other.