TEXT PROOFREADING APPARATUS AND TEXT PROOFREADING METHOD

• US 20130246048A1
• Filed: 12/11/2012
• Published: 09/19/2013
• Est. Priority Date: 03/19/2012
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: NAGASE, Tomoki, CHENG, Yuchang
• Granted Patent: US 9,087,047 B2
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 40/129 Handling non-Latin characters, e.g. kana-to-kanji conversion
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30

Abstract

A Japanese proofreading apparatus has a correction history corpus, a proofreading candidate generation unit, a proofreading availability determination unit, and an automatic proofreading unit. The correction history corpus stores negative sentences as post-proofreading sentences and positive example sentences as post-proofreading sentences, in association with each other. The proofreading candidate generation unit acquires the post-proofreading sentences corresponding to the pre-proofreading sentences from the correction history corpus, according to characteristics of a proofreading target sentence. The proofreading availability determination unit selects, from the post-proofreading sentences acquired by the proofreading candidate generation unit, post-proofreading sentences with degrees of similarity between the proofreading target sentence and the post-proofreading sentences equal to or more than a predetermined threshold value, as proofreading candidates. The automatic proofreading unit proofreads the proofreading target sentence, using, out of the post-proofreading sentences selected by the proofreading availability determination unit, a post-proofreading sentence with the highest degree of similarity.

Claims

1. A text proofreading apparatus comprising:

a storage unit that stores pre-proofreading sentences and post-proofreading sentences in association with each other;

an acquisition unit that acquires the post-proofreading sentences corresponding to the pre-proofreading sentences, according to characteristics of a proofreading target sentence from the storage unit;

a selection unit that selects, from the post-proofreading sentences acquired by the acquisition unit, post-proofreading sentences with degrees of similarity between the proofreading target sentence and the post-proofreading sentences equal to or more than a predetermined threshold value; and

a proofreading unit that proofreads the proofreading target sentence, using, out of the post-proofreading sentences selected by the selection unit, a post-proofreading sentence with the highest degree of similarity.

2. The text proofreading apparatus according to claim 1, wherein

the proofreading unit replaces a character string of an overlapping part in the pre-proofreading sentence between a common character string between the proofreading target sentence and the pre-proofreading sentence and a different character string between the pre-proofreading sentence and the post-proofreading sentence, by a character string in the post-proofreading sentence corresponding to the overlapping character string, thereby proofreading the proofreading target sentence.

3. The text proofreading apparatus according to claim 1, wherein

the selection unit determines availability of proofreading by the post-proofreading sentence based on degree of syntactic similarity or degree of semantic similarity between the proofreading target sentence and the post-proofreading sentence.

4. A text proofreading method comprising:

acquiring post-proofreading sentences corresponding to pre-proofreading sentences, according to a proofreading target sentence;

selecting, from the acquired post-proofreading sentences, post-proofreading sentences with degrees of similarity between the proofreading target sentence and the post-proofreading sentences equal to or more than a predetermined threshold value; and

proofreading the proofreading target sentence, using, out of the selected post-proofreading sentences, a post-proofreading sentence with the highest degree of similarity.