Machine translation apparatus, translation method, and translation system

• US 9,547,645 B2
• Filed: 01/15/2015
• Issued: 01/17/2017
• Est. Priority Date: 01/22/2014
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Kadotani, Masataka, Nagase, Tomoki
• Primary Examiner(s): PULLIAS, JESSE SCOTT
• Application Number:
  US14/597,308
• Publication Number:
  US 20150205788A1
• Time in Patent Office: 733 Days
• Field of Search: 704 2- 8
• US Class Current: 1/1
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/55 Rule-based translation
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation

Abstract

A machine translation apparatus configured to translate an input sentence and output a translated sentence in a target language, the machine translation apparatus includes a rule acquirer configured to acquire a difference between an input example sentence and a replaced example sentence which is obtained by replacing the input example sentence, and acquire a replacement rule based on the difference and each of meaning representations which indicate each relationship of words in the input example sentence and the replaced example sentence; and a translator configured to apply the replacement rule acquired by the rule acquirer to the input sentence and output the translated sentence based on the meaning representations and a meaning representation indicating of relationship of words in the input sentence.

Claims

1. A machine translation apparatus comprising:

a memory and;

at least one processor coupled to the memory, the at least one processor configured to;

obtain a first translated sentence generated by translating a first input sentence and a second translated sentence generated by translating a second input sentence, the second input sentence being a sentence into which the first input sentence is rewritten;

acquire a difference between the first translated sentence and the second translated sentence;

specify a first word in the first translated sentence corresponding to the difference, a second word in the second translated sentence corresponding to the difference and a related word that has a predetermined relationship with the first word and the second word, the related word being included in both of the first translated sentence and the second translated sentence;

acquire a first appearance frequency and a second appearance frequency, the first appearance frequency indicating an appearance frequency of a combination of the first word and the related word with the predetermined relationship in a plurality of example sentences, the second appearance frequency indicating an appearance frequency of a combination of the second word and the related word with the predetermined relationship in the plurality of example sentences;

generate, when the second appearance frequency is larger by a predetermined value than the first appearance frequency, a translation rule for replacing the first word with the second word;

translate an input sentence entered by a user based on the generated translation rule; and

output the translated input sentence.

2. The machine translation apparatus according to claim 1,

wherein when messages are translated, transmitted and received, the translation rule is updated using a feedback re resenting a meaning of the translated sentence and provided by another user who transmits and receives messages.

3. The machine translation apparatus according to claim 1,

wherein the translation rule is performed based on determination of whether or not a node that has a specific modification relation with the difference.

4. The machine translation apparatus according to claim 1,

wherein the translation rule is data set as a combination of a first word to be replaced, a second word to be used to determine that the replacing is to be executed, and a modification relation between the first word and the second word.

5. The machine translation apparatus according to claim 1,

wherein translation from Japanese to English is executed.

6. The machine translation apparatus according to claim 1,

wherein translation from English to Japanese is executed.

7. The machine translation apparatus according to claim 1, wherein

the second input sentence being a sentence into which the first input sentence is rewritten so that the second translated sentence is a more accurate translation than the first translated sentence.

8. A machine translation apparatus configured to translate an input sentence and output a translated sentence in a target language, the machine translation apparatus comprising:

a memory and;

at least one processor coupled to the memory, the at least one processor configured to;

obtain a first translated sentence generated by translating a first sentence and obtain a second translated sentence generated by translating a second sentence, the second sentence being rewritten from the first sentence, the first sentence and the second sentence being acquired from a server apparatus;

acquire a difference between the first translated sentence and the second translated sentence;

specify a first word in the first translated sentence corresponding to the difference, a second word in the second translated sentence corresponding to the difference and a related word that has a predetermined relationship with the first word and the second word, the related word being included in both of the first translated sentence and the second translated sentence;

acquire a first appearance frequency and a second appearance frequency, the first appearance frequency indicating an appearance frequency of a combination of the first word and the related word with the predetermined relationship in a plurality of example sentences, the second appearance frequency indicating an appearance frequency of a combination of the second word and the related word with the predetermined relationship in the plurality of example sentences;

generate, when the second appearance frequency is larger by a predetermined value than the first appearance frequency, a translation rule for replacing the first word with the second word; and

translate an input sentence based on the generated translation rule and output the translated input sentence, wherein

the server apparatus is configured to enable at least two users to exchange messages through a network, the messages being translated by the machine translation apparatus and transmitted and received, the server apparatus enabling the machine translation apparatus to acquire the translation rule based on a feedback provided by any of the users.

9. A translation method comprising:

causing the computer to obtain a first translated sentence generated by translating a first input sentence and a second translated sentence generated by translating a second input sentence, the second input sentence being a sentence into which the first input sentence is rewritten;

causing the computer to acquire a difference between the first translated sentence and the second translated sentence;

causing the computer to specify a first word in the first translated sentence corresponding to the difference, a second word in the second translated sentence corresponding to the difference and a related word that has a predetermined relationship with the first word and the second word, the related word being included in both of the first translated sentence and the second translated sentence;

causing the computer to acquire a first appearance frequency and a second appearance frequency, the first appearance frequency indicating an appearance frequency of a combination of the first word and the related word with the predetermined relationship in a plurality of example sentences, the second appearance frequency indicating an appearance frequency of a combination of the second word and the related word with the predetermined relationship in the plurality of example sentences;

causing the computer to generate, when the second appearance frequency is larger by a predetermined value than the first appearance frequency, a translation rule for replacing the first word with the second word;

causing the computer to translate an input sentence entered by a user based on the generated translation rule; and

causing the computer to output the translated input sentence.