Language analyzer for morphemically and syntactically analyzing natural languages by using block analysis and composite morphemes
First Claim
1. A language analyzer for morphemically and syntactically analyzing natural languages used in an automatic translator, comprising:
- memory means for containing therein morpheme data of a predetermined languages and additional data representing information to be referred to in order that a plurality of morphemes are combined with each other;
first analysis means for morphemically analyzing an inputted character array representing sentences of said language by referring to said morpheme data to break said inputted character array into morphemes and to define respective aspects of said morphemes, havingmeans for distinguishing a part of said inputted character array defining a composite morpheme or a block composed of a plurality of morphemes from the other parts of said inputted character array by referring to said additional data and said morpheme data;
means for treating each of said composite morpheme and said block as a single unit and such that they are regarded as a single morpheme, at the time of syntax analysis, for an inputted character array and for outputting identification as to said one unit;
exclusion means for creating an excluded block containing composite morpheme data or a block which was part of said input string, in response to output from said distinguishing means; and
second analysis means for syntactically analyzing said inputted character array by applying syntactic rules to an analysis result of said first analysis means to describe structures of said sentences, havingmeans for analyzing said part of character array defining the block in preference to analysis for other parts of character array being outside of said excluded block, andmeans for analyzing said inputted character array after analyzing said part of character array by regarding said excluded block as a single morpheme without applying said syntax rules to a relation between a morpheme outside of said excluded block and a morpheme in said excluded block to reduce the number of improper solutions of syntax analysis.
1 Assignment
0 Petitions
Accused Products
Abstract
A language analyzer includes a dictionary having stored therein dictionary data including morpheme data for words, compound words and phrases, and a parsing analyzer for conducting morphological analysis for an inputted sentence by referring to the dictionary. The dictionary contain data for the coupling degree indicating the coupling degree between each of words constituting the compound words or phrases and the parsing analyzer refers to the dictionary for the respective words contained in said inputted sentence and, when a plurality of dictionary data are retrieved for one word in combination with other words, selects the combination of words of a higher coupling degree by referring to the data for said coupling degree.
-
Citations
22 Claims
-
1. A language analyzer for morphemically and syntactically analyzing natural languages used in an automatic translator, comprising:
-
memory means for containing therein morpheme data of a predetermined languages and additional data representing information to be referred to in order that a plurality of morphemes are combined with each other; first analysis means for morphemically analyzing an inputted character array representing sentences of said language by referring to said morpheme data to break said inputted character array into morphemes and to define respective aspects of said morphemes, having means for distinguishing a part of said inputted character array defining a composite morpheme or a block composed of a plurality of morphemes from the other parts of said inputted character array by referring to said additional data and said morpheme data; means for treating each of said composite morpheme and said block as a single unit and such that they are regarded as a single morpheme, at the time of syntax analysis, for an inputted character array and for outputting identification as to said one unit; exclusion means for creating an excluded block containing composite morpheme data or a block which was part of said input string, in response to output from said distinguishing means; and second analysis means for syntactically analyzing said inputted character array by applying syntactic rules to an analysis result of said first analysis means to describe structures of said sentences, having means for analyzing said part of character array defining the block in preference to analysis for other parts of character array being outside of said excluded block, and means for analyzing said inputted character array after analyzing said part of character array by regarding said excluded block as a single morpheme without applying said syntax rules to a relation between a morpheme outside of said excluded block and a morpheme in said excluded block to reduce the number of improper solutions of syntax analysis. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22)
-
-
21. A language analyzer according to claim 86, wherein said second analysis means analyzes an inputted character array, except a portion thereof distinguished as a block whose syntactic attribute and role are estimated at an affirmative or a negative sentence and a tag question respectively by said first analysis means.
Specification