Method and system for topical segmentation, segment significance and segment function

A computer based method for automatically extracting significant topical information from a document, comprising:

• extracting topical information from a document in accordance with specified categories of information;
• linking occurrences of said extracted topical information based on the proximity of similar topical information;
• assigning weighted scores to paragraphs of said document input corresponding to said linked occurrences, wherein said scores depend upon the type of said selected terms and the position of said linked occurrences with respect to said paragraphs, and wherein said scores represent boundaries of said topical segments;
• zero-sum normalizing said assigned weighted scores to determine said topical boundaries;
• determining topical segments within said document corresponding to said linked occurrences of said topical information;
• and
• determining the significance of said topical segments.

A computer program for identifying topical segments of a document input, comprising:

• means for extracting selected terms from a document;
• means for linking occurrences of said extracted terms based upon the proximity of similar terms;
• means for assigning weighted scores to paragraphs of said document input corresponding to said linked occurrences, wherein said scores depend upon the type of said selected terms and the position of said linked occurrences with respect to said paragraphs, and wherein said scores represent boundaries for said topical segments;
• and
• means for zero sum normalizing said assigned weighted scores to determine said topical boundaries.

A computer program for automatically extracting significant topical information from a document, comprising:

• means for extracting topical information from a document in accordance with specified categories of information;
• means for linking occurrences of said extracted topical information based on the proximity of similar topical information;
• means for assigning weighted scores to paragraphs of said document input corresponding to said linked occurrences, wherein said scores depend upon the type of said selected terms and the position of said linked occurrences with respect to said paragraphs, and wherein said scores represent boundaries of said topical segments;
• means for determining topical segments within said document corresponding to said linked occurrences of said topical information;
• and
• means for determining the significance of said topical segments including determining a segment importance and determining a segment coverage.