Language processing application system with status data sharing among language processing functions

• US 5,765,179 A
• Filed: 06/07/1995
• Issued: 06/09/1998
• Est. Priority Date: 08/26/1994
• Status: Expired due to Term

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Kumano, Akira, Sumita, Kazuo, Kanazawa, Hiroshi
• Primary Examiner(s): Jankus, Almis R.
• Application Number:
  US08/475,688
• Time in Patent Office: 1,098 Days
• Field of Search: 395/751-760, 395/792-798, 707/530-536
• US Class Current: 715/236
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/53 Processing of non-Latin text kana-to-kanji conversion G06F40/129; vowelisation G06F40/232
• 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 language processing application system capable of operating a plurality of language processing functions on different computer systems while improving a performance level of each language processing function. In the system, mutually different language processing functions are provided by a plurality of language processing servers, and status data obtained in a course of a language processing by each language processing server is managed at a status data management unit, where the status data indicates at least a part of features of input data processed by each language processing server. Each language processing server acquires the status data obtained in a course of a language processing by another language processing server from the status data management unit and adjusts its language processing according to the status data acquired from the status data management unit.

Claims

1. A language processing application system, comprising:

a plurality of language processing servers capable of providing mutually different language processing functions; and

status data management means capable of managing status data obtained from a language processing server of said plurality of language processing servers in the course of language processing, the status data indicating at least one of features of input data processed by said language processing server in the course of language processing and being accompanied by information indicating a phase of language processing by the language processing server which provided the status data;

each of said plurality of language processing servers being capable of receiving status data obtained from said language processing server by the status data management means, and each being further capable of adjusting its language processing according to said status data,

the status data management means being capable of broadcasting status data to each of said plurality of language processing servers, and

each language processing server being further capable of determining whether to adjust its language processing according to status data broadcast by the status data management means based on a phase of language processing of a language processing server indicated by the status data.

2. The system of claim 1, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, the speech synthesis server being capable of adjusting a sound level of speech to be synthesized in the course of speech synthesis processing according to information regarding a level of a background noise which is obtained by the speech recognition server in the course of a speech recognition processing and which is received by the speech synthesis server as status data from the status data management means.

3. The system of claim 1, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, the speech synthesis server being capable of lowering a sound level of speech to be synthesized in the course of speech synthesis processing according to information indicating a higher level of a speech input which is obtained by the speech recognition server in the course of a speech recognition processing and which is received by the speech synthesis server as status data from the status data management means.

4. The system of claim 1, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, the speech recognition server being capable of giving a higher priority as a recognition target word to each word which is synthesized in the course of speech synthesis processing by the speech synthesis server and which is received by the speech recognition server as status data from the status data management means.

5. The system of claim 1, wherein said plurality of language processing servers includes a speech recognition server and a machine translation server, the speech recognition server being capable of limiting a vocabulary to be recognized in a course of speech recognition processing according to information regarding translation words which is obtained by the machine translation server in the course of machine translation processing and which is received by the speech recognition server as status data from the status data management means.

6. The system of claim 1, wherein said plurality of language processing servers includes a kana-kanji conversion server and a machine translation server, the kana-kanji conversion server being capable of limiting candidate conversion results in the course of kana-kanji conversion processing according to information regarding translation words which is obtained by the machine translation server in the course of machine translation processing and which is received by the kana-kanji conversion server as status data from the status data management means.

7. The system of claim 1, wherein said plurality of language processing servers includes a full text retrieval server and a character recognition server, the character recognition server being capable of switching a recognition dictionary to be used in the course of character recognition processing according to information regarding a subject field of documents stored in a document database for use in a course of full text retrieval processing by the full text retrieval server which is received by the character recognition server as status data from the status data management means.

8. The system of claim 1, wherein said plurality of language processing servers includes a full text retrieval server, a machine translation server, and a character recognition server, the machine translation server being capable of switching a translation dictionary to be used in the course of a machine translation processing and the character recognition server being capable of switching a recognition dictionary to be used in the course of character recognition processing according to information regarding a subject field of documents stored in a document database for use in the course of full text retrieval processing by the full text retrieval server which is received by the machine translation server and the character recognition server as status data from the status data management.

9. The system of claim 1, wherein the status data management means is capable of arbitrating conflict among conflicting status data provided by said plurality of language processing servers.

10. The system of claim 9, wherein the status data management means is capable of arbitrating the conflict by deleting status data having a low priority level among said conflicting status data.

11. The system of claim 10, wherein the status data management means is capable of determining a priority level of each conflicting status data according to a predetermined priority order.

12. The system of claim 10, wherein the status data management means is capable of determining a priority level of each conflicting status data according to a certainty level of each conflicting status data determined in the course of language processing by a language processing server.

13. The system of claim 1, wherein the status data indicating at least one of features of input data processed by a language processing server indicates said at least one feature numerically or symbolically.

14. A language processing application method, comprising:

providing, in each of a plurality of language processing servers, one of mutually different language processing functions;

managing, in a status data managing means, status data obtained in the course of language processing by a first of said plurality of language processing servers, the status data indicating at least one of features of input data processed by said first language processing server and being accompanied by information indicating a phase of language processing of said first language processing server;

acquiring, in a second of said plurality of language processing servers, status data obtained by said status data managing means from said first language processing server;

adjusting language processing in said second language processing server according to said status data obtained by said status data management means;

broadcasting status data received from said first language processing server to each of said plurality of language processing servers; and

determining in each language processing server whether to adjust language processing according to the status data broadcast by the status data management means, based on a phase of language processing of said first language processing server.

15. The method of claim 14, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, and the speech synthesis server adjusts a sound level of speech to be synthesized in the course of speech synthesis processing according to information regarding a level of background noise which is obtained in the course of speech recognition processing by the speech recognition server and which is acquired as status data from the status data management means.

16. The method of claim 14, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, and the speech synthesis server lowers a sound level of speech to be synthesized in the course of speech synthesis processing according to information indicating a higher level of speech input which is obtained in the course of speech recognition processing by the speech recognition server and which is acquired as status data from the status data management means.

17. The method of claim 14, wherein said plurality of language processing servers includes a speech recognition server and a speech synthesis server, and the speech recognition server gives a higher priority as a recognition target word to each word which is synthesized in the course of speech synthesis processing by the speech synthesis server and which is acquired as status data from the status data management means.

18. The method of claim 14, wherein said plurality of language processing servers includes a speech recognition server and a machine translation server, and the speech recognition server limits a vocabulary to be recognized in the course of speech recognition processing according to information regarding translation words which is obtained in the course of machine translation processing by the machine translation server and which is acquired as status data from the status data management means.

19. The method of claim 14, wherein said plurality of language processing servers includes a kana-kanji conversion server and a machine translation server, and the kana-kanji conversion server limits candidate conversion results in the course of kana-kanji conversion processing according to information regarding translation words which is obtained in the course of machine translation processing by the machine translation server and which is acquired as status data from the status data management means.

20. The method of claim 14, wherein said plurality of language processing servers includes a full text retrieval server and a character recognition server, and the character recognition server switches a recognition dictionary to be used in the course of character recognition processing according to information regarding a subject field of documents which is stored in a document database for use in the course of full text retrieval processing by the full text retrieval server and which is acquired as status data from the status data management means.

21. The method of claim 14, wherein said plurality of language processing servers includes a full text retrieval server, a machine translation server, and a character recognition server, and the machine translation server switches a translation dictionary to be used in the course of machine translation processing while the character recognition server switches a recognition dictionary to be used in the course of character recognition processing according to information regarding a subject field of documents which is stored in a document database for use in the course of full text retrieval processing by the full text retrieval server and which is acquired from the status data management means as the status data.

22. The method of claim 14, further comprising arbitrating in the status data management means a conflict among conflicting status data provided by said plurality of language processing servers.

23. The method of claim 22, wherein said arbitrating comprises deleting status data having a low priority level among said conflicting status data.

24. The method of claim 23, wherein a priority level of status data is determined according to a predetermined priority order.

25. The method of claim 23, wherein a priority level of status data is determined according to a certainty level of status data determined in the course of language processing by a language processing server.

26. The method of claim 14, wherein status data indicating at least one of features of input data processed by a language processing server represents said at least one of features numerically or symbolically.