Content availability for natural language processing tasks

• US 9,792,276 B2
• Filed: 12/13/2013
• Issued: 10/17/2017
• Est. Priority Date: 12/13/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Eisen, Jeffrey N.
• Primary Examiner(s): He, Jialong
• Application Number:
  US14/106,324
• Publication Number:
  US 20150169545A1
• Time in Patent Office: 1,404 Days
• Field of Search: 704 1- 9, 704 10, 704258-269, 704270, 704E13001-E13014
• CPC Class Codes: G06F 40/103 Formatting, i.e. changing of presentation of documents automatic justification G06F40/189; automatic line break hyphenation G06F40/191
• CPC Class Codes: G06F 40/151 Transformation
• CPC Class Codes: G06F 40/166 Editing, e.g. inserting or deleting
• CPC Class Codes: G06F 40/169 Annotation, e.g. comment data or footnotes
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

An approach is provided to make content available to natural language processing (NLP) tasks. In the approach, a screen view of a document section is provided as input to a screen reader application. The screen reader application converts information displayed on the screen into a natural language format. A NLP operation is then performed on the natural language format.

Claims

1. A method, in an information handling system comprising a processor and a memory, of making content available to natural language processing (NLP) tasks, the method comprising:

determining that a document section comprises non-textual image data;

in response to the determining, converting the document section into a natural language format, wherein the natural language format comprises textual data compatible with the NLP tasks, and wherein the converting comprises;

selecting a screen reader application, wherein the selected screen reader application does not provide textual output;

inputting a screen view of the document section into the selected screen reader application;

in response to the inputting, receiving, from the selected screen reader application, an audible speech output; and

inputting the audible speech output as an audible speech input to a speech recognition application, wherein the speech recognition application converts the audible speech input into a natural language textual output that describes the non-textual image data; and

performing a NLP operation on the natural language textual output, wherein the NLP operation is carried out by a question and answer system.

2. The method of claim 1 further comprising:

prior to the converting;

receiving a document that includes the document section; and

wherein the determining further comprises identifying that the document section is incompatible with one or more native converters used to convert document contents to the natural language format.

3. The method of claim 1 further comprising:

receiving an input document that includes a plurality of document sections including the document section comprising non-textual image data;

identifying a first set of document sections from the plurality of document sections that are compatible with one or more native converters used to convert document contents to the natural language format;

converting the first set of document sections to the natural language format using the one or more native converters;

identifying a second set of document sections from the plurality of document sections that are incompatible with the one or more native converters, wherein the second set of document sections includes the document section comprising non-textual image data; and

converting the second set of document sections to the natural language format by inputting one or more screen views corresponding to each of the second set of document sections to the selected screen reader application.

4. The method of claim 3 wherein the identifying the second set of document sections further comprises:

detecting a content type corresponding to one or more document sections included in the second set of document sections, wherein the detected content type is selected from a group consisting of a diagram, a table, a picture, and a non-textual document type.

5. The method of claim 3 further comprising:

merging the converted first set of document sections and the converted second set of document sections to a natural language output file, wherein a merge order of the first and second set of converted document sections in the natural language output file is consistent with an original order of the plurality of document sections found in the input document.

6. An information handling system comprising:

one or more processors;

a memory coupled to at least one of the processors;

a display; and

a set of instructions stored in the memory and executed by at least one of the processors to make content available to natural language processing (NLP) tasks, wherein the set of instructions perform actions of;

determining that a document section comprises non-textual image data;

in response to the determining, converting the document section into a natural language format, wherein the natural language format comprises textual data compatible with the NLP tasks, and wherein the converting comprises;

selecting a screen reader application, wherein the selected screen reader application does not provide textual output;

inputting a screen view of the document section into the selected screen reader application;

in response to the inputting, receiving, from the selected screen reader application, an audible speech output; and

inputting the audible speech output as an audible speech input to a speech recognition application, wherein the speech recognition program converts the audible speech input into a natural language textual output that describes the non-textual image data; and

performing a NLP operation on the natural language textual output, wherein the NLP operation is carried out by a question and answer system.

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

prior to the converting;

receiving a document that includes the document section; and

wherein the determining further comprises identifying that the document section is incompatible with one or more native converters used to convert document contents to the natural language format.

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

receiving an input document that includes a plurality of document sections including the document section comprising non-textual image data;

identifying a first set of document sections from the plurality of document sections that are compatible with one or more native converters used to convert document contents to the natural language format;

converting the first set of document sections to the natural language format using the one or more native converters;

identifying a second set of document sections from the plurality of document sections that are incompatible with the one or more native converters, wherein the second set of document sections includes the document section comprising non-textual image data; and

converting the second set of document sections to the natural language format by inputting one or more screen views corresponding to each of the second set of document sections to the selected screen reader application.

9. The information handling system of claim 8 wherein the identifying the second set of document sections further comprises:

detecting a content type corresponding one or more document sections included in the second set of document sections, wherein the detected content type is selected from a group consisting of a diagram, a table, a picture, and a non-textual document type.

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

merging the converted first set of document sections and the converted second set of document sections to a natural language output file, wherein a merge order of the first and second set of converted document sections in the natural language output file is consistent with an original order of the plurality of document sections found in the input document.

11. A non-transitory computer readable storage medium, comprising computer instructions stored thereon, that, when executed by an information handling system, causes the information handling system to make content available to natural language processing (NLP) tasks by performing actions comprising:

determining that a document section comprises non-textual image data;

in response to the determining, converting the document section into a natural language format, wherein the natural language format comprises textual data compatible with the NLP tasks, and wherein the converting comprises;

selecting a screen reader application, wherein the selected screen reader application does not provide textual output;

inputting a screen view of the document section into the selected screen reader application;

in response to the inputting, receiving, from the selected screen reader application, an audible speech output; and

inputting the audible speech output as an audible speech input to a speech recognition application, wherein the speech recognition application converts the audible speech input into a natural language textual output that describes the non-textual image data; and

performing a NLP operation on the natural language textual output, wherein the NLP operation is carried out by a question and answer system.

12. The non-transitory computer readable storage medium of claim 11 wherein the actions further comprise:

prior to the converting;

receiving a document that includes the document section; and

wherein the determining further comprises identifying that the document section is incompatible with one or more native converters used to convert document contents to the natural language format.

13. The non-transitory computer readable storage medium of claim 11 wherein the actions further comprise:

receiving an input document that includes a plurality of document sections including the document section comprising non-textual image data;

identifying a first set of document sections from the plurality of document sections that are compatible with one or more native converters used to convert document contents to the natural language format;

converting the first set of document sections to the natural language format using the one or more native converters;

identifying a second set of document sections from the plurality of document sections that are incompatible with the one or more native converters, wherein the second set of document sections includes the document section comprising non-textual image data; and

converting the second set of document sections to the natural language format by inputting one or more screen views corresponding to each of the second set of document sections to the selected screen reader application.

14. The non-transitory computer readable storage medium of claim 13 wherein the identifying the second set of document sections further comprises:

detecting a content type corresponding to one or more document sections included in the second set of document sections, wherein the detected content type is selected from a group consisting of a diagram, a table, a picture, and a non-textual document type; and

wherein the actions further comprise;

merging the converted first set of document sections and the converted second set of document sections to a natural language output file, wherein a merge order of the first and second set of converted document sections in the natural language output file is consistent with an original order of the plurality of document sections found in the input document.