Abstract: A machine translation system capable of obtaining a consistent translation for an entire document by taking context into account in translating each word or sentence. In this system, document information necessary to remove uncertainty in the translation due to a presence of a plurality of candidates for the translation is utilized whenever uncertainity due to a presence of a plurality of candidates for the translation arises by attempting to translate according to a translation dictionary containing rules for translation.

Abstract: A translation communication system having memory means for storing past communication information and past translated information including as original word/translated word pair obtained from the communication information, and supply means for supplying the translated information useful for determining the translated word candidate to the translation means. With this construction, when the translation means translates a first language into a second language in conversation, the supply means retrieves the communication information in the first language from the memory means in accordance with the input original sentence in the first language, and supplies an original word/translated word pair in accordance with the retrieval result. Therefore, as the translated word in the second language, the present system uses the same word as the original word previously input by the other side's operator who uses the second language.

Abstract: A machine translation system having input means having different function keys, edit keys, cursor keys as well as having normal character keys and numeric keys, for performing various functions for translation, edit control means for controlling the edition operation of the system, translation means coupled to the edit control means, for executing the translation processing in accordance with control signals from the edit control means, and dictionary means having at least a first dictionary for storing various words in various parts of speech in a first language and their translation in a second language corresponding to the words in the first language, and a second dictionary for storing various terminology in noun in the second language corresponding to the words in the first language as well as storing various grammars for the analysis of the languages with this constriction, the meaning of particular kinds of the parts of speech other than a noun takes precedence of the meaning of the Noun when determini

Abstract: An automatic syntax analyzing method is applied to a natural language processing system. It includes the step of detecting by dictionary consultation the syntactic category of words forming an inputted sentence and steps of assumptively applying, when the input sentence includes at least a word unregistered in the dictionary section, an adequate category which satisfies a predetermined grammatical regulation of the input sentence to the unregistered word independent of the dictionary consultation.