Principle of Operation
Summarizer performs a semantic analysis of the text. It extracts Verb-Object pairs and defines the relationship between them, thus forming concepts.
Summarizer determines the semantic weight of each concept and creates the summary accordingly. If the summary type is specified, Summarizer will place primary importance upon specific concepts and structure, thus making the summary theme- and structure-oriented.
If the user finds some concepts more important than other he/she can check them in the concept tree. Semantic weight of the checked concepts gets increased, and as a result sentences that comprise these concepts are ranked higher and have greater probability to be included into a summary.
In addition to generating a summary, Summarizer forms a concept tree. Its branched structure depicts the text's logic and the semantic weight of each key concept.
Summaries can be saved to a "History" in order to keep track of the generated summaries. For the purpose of convenience, these can be sorted by date, type or concept.