Like the Brainware Classify Engine, the ASSA Classify Engine needs a learnset of documents for classification. The difference between the Brainware Engine and the ASSA Engine lies in the different search pools that are generated out of the documents of the learnset when the documents are learned.
For its search pool, Brainware Classify Engine generates a dictionary of words. The documents for the search pool are coded with words from this dictionary.
The ASSA engine uses the trigram method to code the search pool. The classification of documents bases on the block text search, which involves retrieving the text from the document block by block and generating a query string out of the block text. The ASSA engine then searches for that query within the search pool.