Lexical correction of erroneous text by transformation into a voice message

• US 8,249,869 B2
• Filed: 06/15/2007
• Issued: 08/21/2012
• Est. Priority Date: 06/16/2006
• Status: Expired due to Fees

• Current Assignee: Abderrahman Brini
• Original Assignee: Logolexie Logolexie
• Inventors: Vessiere, Gilles, Bachelerie, Joël
• Primary Examiner(s): Lerner, Martin
• Application Number:
  US12/308,378
• Publication Number:
  US 20100161312A1
• Time in Patent Office: 1,894 Days
• Field of Search: 704/231, 704/234, 704/235, 704/246, 704/250, 704/258, 704/260, 704/261, 704/266, 704/270
• US Class Current: 704/235
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation

Abstract

The method is suitable for dysorthographic or partially sighted persons, to facilitate the semantic, syntactic and/or lexical correction of an erroneous expression in a digital text input by a user. The method comprises the sequence of: a step (74) of transforming the digital text into a digital voice message, in which the graphemes of the erroneous textual expression are converted into phoneme(s) constituting an intelligible vocal expression, then a step (78) of processing the digital voice message obtained at the end of the transformation step (74), in which the phoneme or phonemes constituting the intelligible vocal expression are converted into grapheme(s) constituting a corrected textual expression of the erroneous textual expression, with the aid of pre-established writing rules.

Claims

1. A method of semantic, syntactic and/or lexical correction of an erroneous expression in a digital text, comprising:

using a computer, transforming a digital text comprising eventual erroneous textual expressions into a digital voice message, in which at least one grapheme of the digital text comprising eventual erroneous textual expressions is converted into at least one phoneme comprising an intelligible vocal expression of the entire digital text, including any erroneous textual expressions,

using the computer and pre-established writing rules, processing the digital voice message obtained at the transformation step, in which the at least one phoneme comprising the intelligible vocal expression is converted into at least one grapheme comprising a new digital text in which any erroneous textual expression is corrected,

during the transformation step, providing the at least one phoneme according to a predetermined voice model, with the aid of a voice model base, and

during, the processing step, identifying the at least one phoneme to be converted, with the aid of a base of recognition rules for the predetermined voice model.

2. The method as claimed in claim 1, further comprising:

before the transformation step, using the computer, providing a digital set of textual information data representative of the digital text with the erroneous expression,

wherein the transformation step comprises converting the set of textual information data into a digital set of vocal information data representative of the at least one phoneme associated with the at least one grapheme of the digital text, including the at least one phoneme comprising the intelligible vocal expression, and

wherein the processing step comprises converting the vocal information data provided by the conversion operation of the transformation step, into a digital set of textual information data representative of a reconstituted digital text including the corrected textual expression.

3. The method of claim 1, further comprising, between the transformation step and the processing step, using the computer for synthesizing the sound of the digital voice message obtained at the transformation step, at the end of which synthesis step the digital voice message to be processed during the processing step is validated.

4. A semantic, syntactic and/or lexical electronic correction apparatus for facilitating the semantic and/or syntactic and/or lexical correction of an erroneous expression in a digital text comprising eventual erroneous textual expressions, comprising:

a module using a computer for transforming the entire digital text into a digital voice message, the transformation module being able to convert at least one grapheme of the digital text comprising eventual erroneous textual expressions into at least one phoneme comprising an intelligible vocal expression of the entire digital text, including any erroneous textual expression, and

a module using the computer and a pre-established base of writing rules for processing the digital voice message provided by the transformation module, the processing module being able to convert the at least one phoneme comprising the intelligible vocal expression and provided by the transformation module, into at least one grapheme comprising a new digital text in which any erroneous textual expression is corrected,

wherein the transformation module uses the computer to provide the at least one phoneme according to a predetermined voice model, with the aid of a voice model base, and

wherein the processing module is able to identify the at least one phoneme to be converted, with the aid of a base of recognition rules for the predetermined voice model.

5. The electronic correction apparatus of claim 4, further comprising:

at least one loudspeaker able to synthesize as audible sounds the digital voice message provided by the transformation module, and

an activation device for activating the processing module so as to apply the processing module to the digital voice message synthesized by the loudspeaker.

6. The electronic correction apparatus of claim 4, wherein the transformation module uses the computer to segment the digital voice message into at least one word, with the aid of a pre-established base of segmentation rules.

7. The electronic correction apparatus of claim 4, further including a module using a computer for automatic orthographic correction of the corrected textual expression provided by the processing module.

8. A computer program embodied in an information recording medium, comprising instructions for executing a correction method when the instructions are executed by a computer, the method comprising:

using the computer, transforming a digital text comprising eventual erroneous textual expressions into a digital voice message, in which at least one grapheme of the digital text comprising eventual erroneous textual expressions is converted into at least one phoneme comprising an intelligible vocal expression of the entire digital text, including any erroneous textual expression, and

using the computer and pre-established writing rules;

processing the digital voice message obtained at the transformation step, in which the at least one phoneme comprising the intelligible vocal expression is converted into at least one grapheme comprising a new digital text in which any erroneous textual expression is corrected,

during the transformation step, providing the at least one phoneme according to a predetermined voice model with the aid of a voice model base, and

during the processing step, identifying the at least one phoneme to be converted, with the aid of a base of recognition rules for the predetermined voice model.