ABSTRACT
Understanding a discourse is considered to involve a series of specific processing phases which final result is a complete semantic, mental representation (Johnson-Laird, 1983; van Dijk & Kintsch, 1983). This result is not only a representation of the text, but rather of what the text is about. When a reader is asked to summarize a discourse, vast amounts of information within the discourse are selectively ignored in order to produce a distilled version of the original text. This simplification process emphasizes central elements of the discourse while the peripheral details are neglected. It is further demonstrated that discourse can be represented as a skeleton in which the relationships among the clauses could be chunked in a way that replicated the semantic structure of the original discourse (Grosz & Sidner, 1986). Textual continuity, which differentiates a text from a random sequence of sentences, is a prime factor in discourse summarization (Ehrlich & Charolles, 1991).
