System for automatic language translation using several dictionary storage areas and a noun table

• US 5,020,021 A
• Filed: 01/10/1986
• Issued: 05/28/1991
• Est. Priority Date: 01/14/1985
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Kaji, Hiroyuki, Nitta, Yoshihiko
• Primary Examiner(s): Williams, Jr., Archie E.
• Assistant Examiner(s): Harrell, Robert B.
• Application Number:
  US06/817,624
• Time in Patent Office: 1,964 Days
• Field of Search: 364/900 MS File, 364/200 MS File, 364/419, 364/300
• US Class Current: 704/5
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/55 Rule-based translation

Abstract

A translation method for a machine translation system provided with apparatus for parsing a source language sentence and for forming a target language translation in which a phrase omitted in the source language setence is identified, and a word or phrase to be inserted for the omitted phrase is selected from stored words and phrases. For identifying an omitted phrase, a sentence pattern corresponding to a predicate in the source language sentence is formed so as to include not only cases governed by the predicate but also a semantic feature for each case. By comparing the source language sentence with the sentence pattern, a case which is omitted in the source language sentence but cannot be omitted in the target language translation is identified. For determining a word or phrase to be placed at the position of the omitted phrase, the nouns having appeared in the source language text is stored in a noun, together with the semantic feature, gender, person and number of each noun is searched for a noun having the same semantic feature as the omitted phrase. When a target language translation of the source language sentence is formed, a pronoun having the same gender, person and number as the omitted phrase is used as a target language equivalent for the omitted phrase, and thus a target language translation which is grammatically correct, is obtained.

Claims

1. A translation method for a machine translation apparatus for translating a source language sentence being input to said machine translation apparatus into an equivalent target language sentence, said source language sentence including a plurality of words and being inputted to said machine translation apparatus in a form of an electrical signal from an input device, said machine translation apparatus carrying out predetermined translation processes by referring to an electronic data storage device having a word dictionary storage area, a sentence pattern dictionary storage area and a table storage, said word dictionary storage area storing a plurality of word records each defining a relation among a source language word, a part of speech of said source language word, a target language word corresponding to said source language word and a semantic feature of these words, said sentence pattern dictionary storage area including a plurality of sentence pattern records each of which defines cases governed by a predicated in the source language sentence and includes semantic features of respective cases, said cases being arranged in each sentence pattern record in accordance with a word order of target language words in an apparent structure of the target language sentence, said method comprising the steps of:

(a) dividing said source language sentence inputted by said input device into a plurality of words by referring to said word dictionary storage area to search for a predicate and nouns in said source language sentence;

(b) storing contents of respective word records of said word dictionary storage area, each of which includes a source language word corresponding to one of the nouns obtained by a search of said input source language sentence, sequentially into a noun stack area within said table storage area;

(c) parsing said source language sentence by referring to one sentence pattern record in said sentence pattern dictionary storage area which includes a predicate same as that obtained by the search of said input source language sentence, and judging whether a case, which is indispensable for the translating of the input source language sentence into a target language sentence, is omitted in said input source language sentence or not;

(d) when it is judged at the step (c) that a case is omitted, searching, among word records in said noun stack area, for one word record having the same semantic feature as that of said omitted case of said input source language sentence; and

(e) substituting a noun defined by said one word record obtained by the search of said noun stack area for said omitted case, and arranging a target language word for said noun to be substituted and target language words corresponding to words of said input source language sentence obtained from word records in said word dictionary storage area in accordance with the word order defined by said one sentence pattern record.

2. A translation method according to claim 1, wherein

said electronic data storage device has a personal pronoun table storage area for storing declensions of personal pronouns,

each of the word records stored in said word dictionary storage area includes a gender and a person of a target language word,

each of the word records stored in said noun stack area in said step (b) includes a semantic feature, gender and a person for the noun, and

said step (e) supplies said omitted case with one pronoun which is obtained based on a content of said one word record obtained by the search of said noun stack area and a content of said personal pronoun table storage area.

3. A translation method according to claim 1, wherein

said step (c) includes a substep of temporarily storing in said table storage area an intermediate expression frame comprising a plurality of cases and target word records, said cases corresponding to cases of said one sentence pattern and having an arrangement same as that of said one sentence pattern, each of said target word records being associated with one of said cases and obtained from one of said word records in said word dictionary storage area which includes source language words corresponding to the words in said input source language sentence,

said step (d) includes a substep of adding to said intermediate expression frame a target word record for said omitted case which was obtained from said one word record searched in said noun stack area, and

said step (e) forms said target language sentence based on said intermediate expression frame stored in said table storage area.

4. A translation method according to claim 3, wherein

said electronic data storage device has a personal pronoun table storage area for storing declensions of personal pronouns,

each of the word records in said word dictionary storage area includes a gender and a person of a target language word,

each of the word records stored in said noun table area in said step (b) includes a semantic feature, gender and person for the noun, and

said step (e) includes a substep of substituting a person for said omitted case based on a content of said personal pronoun table storage area and gender and person in said one word record obtained by the search of said noun stack area as said intermediate expression frame.

5. A translation method according to claim 1, wherein

said noun stack area stores sequentially a plurality of word records, each of which includes information as to a noun corresponding to one of nouns extracted from source language sentences which have been translated already, and

said step (b) is executed after said step (d).

6. A translation method according to claim 5, wherein said search of said noun table area in said step (d) is performed by checking whether each of the word records stored in said noun table area has the same semantic feature as said omitted case in accordance with an order of the word record lastly stored therein.