Reordering text from unstructured sources to intended reading flow

• US 9,658,990 B2
• Filed: 09/18/2014
• Issued: 05/23/2017
• Est. Priority Date: 09/18/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bruno, Nicholas V., Smythe, Jared M.
• Primary Examiner(s): STORK, KYLE R
• Application Number:
  US14/490,076
• Publication Number:
  US 20160085727A1
• Time in Patent Office: 978 Days
• CPC Class Codes: G06F 40/134 Hyperlinking
• CPC Class Codes: G06F 40/151 Transformation

Abstract

An approach is provided in which a number of sections from a sequence of characters included in a Portable Document Format (PDF) file are identified. Each of the identified sections includes a unique set of coordinate positions. The approach builds links between the sections based on a relative position of each of the sections in relation to the other sections along an axis. The approach repeatedly merges sections based on the links that were built to form increasingly larger sections until a final larger section is generated with the characters appearing in a manner consistent with human reading of the rendered PDF document rather than the placement of the characters found within the original PDF file.

Claims

1. An information handling system comprising:

one or more processors;

a memory coupled to at least one of the processors; and

a set of computer program instructions stored in the memory and executed by at least one of the processors in order to perform actions of;

identifying a plurality of sections from a sequence of characters included in a Portable Document Format (PDF) source file, wherein each section includes a unique set of coordinate positions;

building a plurality of directional links between the plurality of sections based on a relative position of each sections'"'"' coordinate positions in relation to other sections'"'"' coordinate positions along an axis; and

repeatedly merging two or more sections to form increasingly larger sections, wherein the merged two or more sections are selected based on the directional links built between the two or more sections, wherein the repeatedly merging further comprises building one or more new directional links between the increasingly larger sections and one or more remaining sections selected from the plurality of sections, and wherein the repeatedly merging continues until the plurality of sections are exhausted and consolidated into a final larger section, wherein the final larger section is arranged in an intended reading order.

2. The information handling system of claim 1 wherein the building of the directional links further comprises:

selecting one of the sections from the plurality of sections;

building zero or more directional links between the selected section and the other sections included in the plurality of sections by;

establishing zero or more vertical links between the selected section from the plurality of sections and a reference section selected from the plurality of sections wherein the selected section has at least one common horizontal coordinate position with the selected reference section and wherein a vertical rectangle space formed by a horizontal boundary of the selected section and a corresponding horizontal boundary the selected reference section is void of any intervening sections from the plurality of sections; and

establishing zero or more horizontal links between the selected section and the selected reference section wherein the selected section has at least one common vertical coordinate position with the selected reference section and wherein a horizontal rectangle space formed by a vertical boundary of the selected section and a corresponding vertical boundary the selected reference section is void of any intervening sections from the plurality of sections; and

repeatedly selecting a next one of the sections from the plurality of sections and building the zero or more directional links until each of the sections from the plurality of sections has been selected.

3. The information handling system of claim 2 wherein the actions further comprise:

specially merging two or more of the plurality of sections by;

identifying a first section at a first set of coordinate positions, wherein the first set of coordinate positions forms a first rectangular boundary around the first section, and wherein the identification of the first section is based on at least one special rule;

identifying a second section based on at a second set of coordinate positions, wherein the second set of coordinate positions forms a second rectangular boundary around the second section, and wherein the identification of the second section is based on a proximity to the first section;

merging the first and second sections to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the first and second sections with the other sections included in the plurality of sections.

4. The information handling system of claim 3 wherein identification of the first section based on one of the special rules further comprises:

identifying that the first section is an island section that is void of any directional links to other sections from the plurality of sections.

5. The information handling system of claim 3 wherein identification of the first section based on one of the special rules further comprises:

identifying that the first section is an initial character rendered in a first font size that is larger than a second font size used to render a plurality of characters included in the second section.

6. The information handling system of claim 2 wherein the actions further comprise:

vertically merging two or more of the plurality of sections by;

identifying one of the plurality of sections as a selected section and one of the plurality of sections as a reference section, wherein the identification is based on the selected section including a first directional link to the reference section in a first vertical direction and the reference section including a second directional link to the selected section in an second vertical direction, wherein the second vertical direction is opposite from the first vertical direction; and

merging the selected section and the reference section to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the selected section and the reference section with the other sections included in the plurality of sections.

7. The information handling system of claim 6 wherein the actions further comprise:

repeatedly performing the vertical merger of sections until exhaustion of selected and reference sections with vertical links is encountered;

upon exhaustion of selected and reference sections with vertical links, horizontally merging two or more of the plurality of sections by;

identifying one of the plurality of sections as a second selected section and one of the plurality of sections as a second reference section, wherein the identification is based on the second selected section including a primary directional link to the reference section in a first horizontal direction and the second reference section including a secondary directional link to the second selected section in an second horizontal direction, wherein the second horizontal direction is opposite from the first horizontal direction; and

merging the second selected section and the second reference section to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the second selected section and the second reference section with the other sections included in the plurality of sections.

8. The information handling system of claim 7 wherein the actions further comprise:

repeatedly performing the special merging, the vertical merging, and the horizontal merging until the increasingly larger section includes all of the plurality of sections consolidated into the final large section;

inputting the final large section to a corpus utilized by a question answering (QA) system; and

increasing a natural language processing (NLP) performance of the QA system by performing NLP operations on the final large section instead of the PDF source file.

9. A computer program product stored in a computer readable storage medium, comprising computer program code that, when executed by an information handling system, causes the information handling system to perform actions comprising:

identifying a plurality of sections from a sequence of characters included in a Portable Document Format (PDF) source file, wherein each section includes a unique set of coordinate positions;

building a plurality of directional links between the plurality of sections based on a relative position of each sections'"'"' coordinate positions in relation to other sections'"'"' coordinate positions along an axis; and

repeatedly merging two or more sections to form increasingly larger sections, wherein the merged two or more sections are selected based on the directional links built between the two or more sections, wherein the repeatedly merging further comprises building one or more new directional links between the increasingly larger sections and one or more remaining sections selected from the plurality of sections, and wherein the repeatedly merging continues until the plurality of sections are exhausted and consolidated into a final larger section, wherein the final larger section is arranged in an intended reading order.

10. The computer program product of claim 9 wherein the building of the directional links further comprises:

selecting one of the sections from the plurality of sections;

building zero or more directional links between the selected section and the other sections included in the plurality of sections by;

establishing zero or more vertical links between the selected section from the plurality of sections and a reference section selected from the plurality of sections wherein the selected section has at least one common horizontal coordinate position with the selected reference section and wherein a vertical rectangle space formed by a horizontal boundary of the selected section and a corresponding horizontal boundary the selected reference section is void of any intervening sections from the plurality of sections; and

establishing zero or more horizontal links between the selected section and the selected reference section wherein the selected section has at least one common vertical coordinate position with the selected reference section and wherein a horizontal rectangle space formed by a vertical boundary of the selected section and a corresponding vertical boundary the selected reference section is void of any intervening sections from the plurality of sections; and

repeatedly selecting a next one of the sections from the plurality of sections and building the zero or more directional links until each of the sections from the plurality of sections has been selected.

11. The computer program product of claim 10 wherein the actions further comprise:

specially merging two or more of the plurality of sections by;

identifying a first section at a first set of coordinate positions, wherein the first set of coordinate positions forms a first rectangular boundary around the first section, and wherein the identification of the first section is based on at least one special rule;

identifying a second section based on at a second set of coordinate positions, wherein the second set of coordinate positions forms a second rectangular boundary around the second section, and wherein the identification of the second section is based on a proximity to the first section;

merging the first and second sections to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the first and second sections with the other sections included in the plurality of sections.

12. The computer program product of claim 11 wherein identification of the first section based on one of the special rules further comprises:

identifying that the first section is an island section that is void of any directional links to other sections from the plurality of sections.

13. The computer program product of claim 11 wherein identification of the first section based on one of the special rules further comprises:

identifying that the first section is an initial character rendered in a first font size that is larger than a second font size used to render a plurality of characters included in the second section.

14. The computer program product of claim 10 wherein the actions further comprise:

vertically merging two or more of the plurality of sections by;

identifying one of the plurality of sections as a selected section and one of the plurality of sections as a reference section, wherein the identification is based on the selected section including a first directional link to the reference section in a first vertical direction and the reference section including a second directional link to the selected section in an second vertical direction, wherein the second vertical direction is opposite from the first vertical direction; and

merging the selected section and the reference section to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the selected section and the reference section with the other sections included in the plurality of sections.

15. The computer program product of claim 14 wherein the actions further comprise:

repeatedly performing the vertical merger of sections until exhaustion of selected and reference sections with vertical links is encountered;

upon exhaustion of selected and reference sections with vertical links, horizontally merging two or more of the plurality of sections by;

identifying one of the plurality of sections as a second selected section and one of the plurality of sections as a second reference section, wherein the identification is based on the second selected section including a primary directional link to the reference section in a first horizontal direction and the second reference section including a secondary directional link to the second selected section in an second horizontal direction, wherein the second horizontal direction is opposite from the first horizontal direction; and

merging the second selected section and the second reference section to form one of the increasingly larger sections; and

repeating the building of the zero or more directional links between the increasingly larger section formed by the merger of the second selected section and the second reference section with the other sections included in the plurality of sections.

16. The computer program product of claim 15 wherein the actions further comprise:

repeatedly performing the special merging, the vertical merging, and the horizontal merging until the increasingly larger section includes all of the plurality of sections consolidated into the final large section;

inputting the final large section to a corpus utilized by a question answering (QA) system; and

increasing a natural language processing (NLP) performance of the QA system by performing NLP operations on the final large section instead of the PDF source file.