Proofreading aid based on closed-class vocabulary

• US 6,081,772 A
• Filed: 03/26/1998
• Issued: 06/27/2000
• Est. Priority Date: 03/26/1998
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Lewis, James R.
• Primary Examiner(s): Thomas, Joseph
• Application Number:
  US09/048,715
• Time in Patent Office: 824 Days
• Field of Search: 704/1, 704/9, 704/10, 704/235, 704/251, 704/257, 704/275, 704/270, 704/272, 707/530, 707/531, 707/532, 707/533, 434/156, 434/167, 434/169, 434/185
• US Class Current: 704/1
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation

Abstract

A method for proofreading text generated by a speech application can include the following steps beginning with the step of finding all closed-class vocabulary words in the text. Subsequently, the method can include the step of automatically highlighting at least some of the closed-class vocabulary words in the text during proofreading. Additionally, the highlighting of the closed-class vocabulary words can include either highlighting all of the closed-class vocabulary words, highlighting the closed-class vocabulary words based on a perplexity filter, or highlighting the closed-class vocabulary words based on a language model enhancement filter.

Claims

1. A method for proofreading text generated by a speech application, consisting of the steps of:

finding all closed-class vocabulary words in said text; and

,

automatically highlighting only said closed-class vocabulary words in said text during proofreading.

2. The method of claim 1, wherein the step of highlighting highlights only all of said closed-class vocabulary words.

3. The method of claim 1, wherein the step of finding is based on a filtering step of said closed-class vocabulary words in said text and the step of highlighting highlights only those of said closed-class vocabulary words returned by said filtering.

4. The method of claim 3, wherein the step of filtering said closed-class vocabulary words is based on perplexity.

5. The method of claim 3, wherein the step of filtering said closed-class vocabulary words is based on language model enhancement.

6. The method of claim 1, wherein the step of highlighting said closed-class vocabulary words is in accordance with at least one of the following rules:

highlighting all of said closed-class vocabulary words;

highlighting said closed-class vocabulary words based on a perplexity filter; and

,

highlighting said closed-class vocabulary words based on a language model enhancement filter.

7. The method of claim 6, wherein said rules are user selectable.

8. The method of claim 1, wherein the step of highlighting said closed-class vocabulary words is in accordance with at least one of a plurality of predetermined rules.

9. The method of claim 8, wherein at least one of said rules identifies a word selection filter.