This process is based on automatic text indexing techniques which have been established in information retrieval area[#!maa1se4!#,#!luh1se4!#,#!pal2se4!#]. The main technique is extracting lexical affinities from the requirement component. The observation of lexical affinities in large textual document has been shown to convey information on both syntactic and semantic levels and provides us with a powerful way of taking context into account[#!sma1se4!#]. Lexical affinities will serve as indices of the component.