Browse Prior Art Database

Extending the Vector Space Model to Include a Reading Level Dimension

IP.com Disclosure Number: IPCOM000104032D
Original Publication Date: 1993-Feb-01
Included in the Prior Art Database: 2005-Mar-18
Document File: 1 page(s) / 43K

Publishing Venue

IBM

Related People

Wyman, B: AUTHOR

Abstract

A method of extending the vector space model of information retrieval to better represent graduated text levels within a subject is disclosed. The vector space of the model is defined to have one additional dimension that quantifies the reading level of the text. In this fashion, hypertext databases containing information about a subject at several levels can be better partitioned according to the depth at which the subject is discussed in each volume. When the vector space model is to be used for a document database containing texts at several reading levels, preallocate a location in the characteristic vector for the analyzed text's reading level. When analyzing a document and building its characteristic vector for use in the vector space model, place a reading level value in this preassigned vector location.

This text was extracted from an ASCII text file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 81% of the total text.

Extending the Vector Space Model to Include a

Reading

Level Dimension

      A method of extending the vector space model of information
retrieval to better represent graduated text levels within a subject
is disclosed.  The vector space of the model is defined to have one
additional dimension that quantifies the reading level of the text.
In this fashion, hypertext databases containing information about a
subject at several levels can be better partitioned according to the
depth at which the subject is discussed in each volume.
     When the vector space model is to be used for a document
database containing texts at several reading levels, preallocate a
location in the characteristic vector for the analyzed text's reading
level.
     When analyzing a document and building its characteristic vector
for use in the vector space model, place a reading level value in
this preassigned vector location.
     Then,  when the matching function is applied to the documents in
the database, the source's reading level will be a component in the
match function and will help to partition the selected set of target
texts so that text levels more closely matching the source's level
will be ranked as matching more closely than those that might be
either much more or less advanced.
     The figure depicts the processing of the vector space model.
The first step 1 is to take the texts and analyze them for
significant words and phrases, with the output being the definition...