Another goal is to identify salient points: concise lists of different topics, if possible in order of importance, adjustable in depth.An important goal is also taxonomy (keywords in the data base): determination of the topics in the documents which are (or should be) of interest to the reader. This is to be followed by classification: the grouping of documents containing different topics, either as defined by the scriptor, or as defined by the information content.
The most valuable help for the hypertextual reader consists of the identification of dependencies of the different topics on each other, especially of unexpected relationships.
Download ebooks on http://www.frenchtheory.com/
See that post with different algorithms in metabole
See the journal French Metablog with today different posts
