Back to results
US 2006/0080309 A1Published application
Article extraction
AssigneeHewlett-Packard Development Company, L.P.
InventorsSherif Yacoub, Jean-Manuel Thong, John Burns
Life of this patent
Estimating the term…
Abstract
An article is extracted from a document using a decision combiner to process a plurality of reading order alternatives. The text flow analysis generates the plurality of reading order alternatives of separate body text regions.
Claims31
A method comprising:
- extracting articles from a document using text flow analysis to generate a plurality of reading order alternatives of separate body text regions;
- and
- using a decision combiner to process the plurality of reading order alternatives.
A method comprising:
- generating a plurality of separate body text regions with a region determination of each page of a document;
- generating a plurality of reading order alternatives of the plurality of separate body text regions with a text flow analysis;
- and
- applying a decision combiner to process the plurality of reading order alternatives to extract a reading order for each article in the document.
1 more element
A system comprising:
- means for generating a plurality of reading order alternatives from a plurality of separate body text regions for each page of a document;
- and
- means for processing the plurality of reading order alternatives to extract a reading order for each article in the document.
A system comprising:
- a text flow analysis to generate a plurality of reading order alternatives from a plurality of separate body text regions for each page of a document;
- and
- a decision combiner to process the plurality of reading order alternatives to extract a reading order for each article in the document.
A computer-readable medium comprising:
- a plurality of separate body text regions generated by a region determination of each page of a document;
- a plurality of reading order alternatives generate by a text flow analysis of the plurality of separate body text regions, wherein each of the plurality of reading order alternatives identifies one or more hypothetical article sets;
- and
- an algorithm to process each of the plurality of reading order alternatives and the one or more hypothetical article sets to generate a reading order of an article in the document.
1 more element
Dependencies are inferred from the claim text.