Efficient storage mechanism for representing term occurrence in unstructured text documents

A method of converting, organizing, and representing in a computer memory a document corpus containing an ordered plurality of documents, for use by a data mining application program requiring occurrence-of-terms data, said representation to be based on terms in a dictionary previously developed for said document corpus and wherein each said term in said dictionary has associated therewith a corresponding unique integer, said method comprising:

• for said document corpus, taking in sequence each said ordered document and developing a first uninterrupted listing of said unique integers to correspond to the occurrence of said dictionary terms in the document corpus;
• and
• developing a second uninterrupted listing for said entire document corpus, containing in sequence the location of each corresponding document in said first uninterrupted listing, wherein said first listing and said second listing are provided as input data for said data mining application program.

An apparatus for organizing and representing in a computer memory a document corpus containing an ordered plurality of documents, for use by a data mining applications program requiring occurrence-of-terms data, said representation to be based on terms in a dictionary previously developed for said document corpus and wherein each said term in said dictionary has associated therewith a corresponding unique integer, said apparatus comprising:

• an integer determiner receiving in sequence each said ordered document of said document corpus and developing a first uninterrupted listing of said unique integers to correspond to the occurrence of said dictionary terms in the document corpus;
• and
• a locator developing a second uninterrupted listing for said entire document corpus containing in sequence the location of each corresponding document in said first uninterrupted listing, wherein said first listing and said second listing are provided as input data for said data mining applications program.

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform a method to organize and represent in a computer memory a document corpus containing an ordered plurality of documents, for use by a data mining algorithm requiring occurrence-of-terms data, said representation to be based on terms in a dictionary previously developed for said document corpus and wherein each said term in said dictionary has associated therewith a corresponding unique integer, said method comprising:

• a first uninterrupted listing of said unique integers to correspond to the occurrence of said dictionary terms in the document corpus;
• and
• a second uninterrupted listing for said entire document corpus containing in sequence the location of each corresponding document in said first uninterrupted listing, wherein said first listing and said second listing are provided as input data for said data mining algorithm.

A data converter for organizing and representing in a computer memory a document corpus containing an ordered plurality of documents, for use by a data mining applications program requiring occurrence-of-terms data, said representation to be based on terms in a dictionary previously developed for said document corpus and wherein each said term in said dictionary has associated therewith a corresponding unique integer, said data converter comprising:

• means for developing a first uninterrupted listing of said unique integers to correspond to the occurrence of said dictionary terms in the document corpus and;
• and
• means for developing a second uninterrupted listing for said entire document corpus containing in sequence the location of each corresponding document in said first uninterrupted listing, wherein said first listing and said second listing are provided as input data for said data mining applications program.