FORMALIZATION OF A NATURAL LANGUAGE

• US 20120101803A1
• Filed: 11/12/2008
• Published: 04/26/2012
• Est. Priority Date: 11/14/2007
• Status: Abandoned Application

• Current Assignee: Ivaylo Popov, Krasimir Nikolaev Popov
• Original Assignee: Ivaylo Popov, Krasimir Nikolaev Popov
• Inventors: Popov, Ivaylo, Popov, Krasimir Nikolaev
• Application Number:
  US12/740,106
• Publication Number:
  US 20120101803A1
• Time in Patent Office: Days
• US Class Current: 704/2
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/56 Natural language generation

Abstract

It is disclosed a method for formalization of a natural language allowing creation of an unambiguous model of a natural language text. It is determined the basic notions for entities that are named by a natural language and for each basic notion it is attached an unique number or name and a description, in addition it is attached a list of words which can name the basic notion for each used natural language. The unambiguous model uses only basic notions. In this way it is possible a machine to interpret the unambiguous model and to input knowledge and data in a base or to make a text generation in another natural language using the unambiguous model. Also it can be generated a text in artificial language such as a program language.

Claims

1. Formalization of a natural language that enables a machine interpretation and generation of a text in natural language by creating a machine model of the text, characterized by creation of an unambiguous model of the text in natural language which can be interpreted in one and only in one way following these steps:

it is using previously determined basis of notions which the humanity uses so that the basis of notions includes all the basic notions which are unique denotations of an entity or action and they are

unique label—

number or word

and they have

description in a natural language,

and they have

for each natural language which is going to be processed using the method, an attached list of words, which name is in the given natural language;

a computer analyses the text in the natural language and as using the basis of notions and in particular the lists of words which name a certain basic notion in the given natural language it finds used basic notions and together with a grammatical and language analysis it makes first unambiguous model of the text in a natural language;

a computer uses the first unambiguous model to generate again the text in the same natural language;

a computer compares the generated text in a natural language from the first unambiguous model to the original text and it marks the differences;

an operator uses a computer program with which he/she can see the basic notions, chosen by the computer and to change them, also he/she can determine relationships and characteristics of the text which the computer has made difficult finding like which parts of speech are, like for a certain action in which tense it is in a complex sentence or when it is about actions in two adjacent sentences, like what exactly a pronoun substitutes, like which part of the speech with which is connected and how;

a computer uses the operator'"'"'s remarks and the first unambiguous model and generates a second unambiguous model;

a computer uses the second unambiguous model to generate again the text in the same natural language;

a computer compares the generated text in a natural language from the second unambiguous model with the original text and it marks the differences;

an operator makes corrections and the steps interpretation-generation-correction are repeated while the operator accepts that the recently generated from the computer unambiguous model presents the meaning of the text in a natural language well enough.

2. Formalization of a natural language, according to claim 1, characterized also by the step where the formed unambiguous model of the text in a natural language is attached to the same text by a link or by putting the file with the text in a natural language together whit the file containing its unambiguous model in one archive package.

3. Formalization of a natural language, according to claim 1, characterized also by the step where the unambiguous model of the text in a natural language is used in machine processes like searching, extracting facts and relationships, also like in deteraiining a text in its legal meaning.

4. Formalization of a natural language, according to claim 1, characterized also by the step where it uses comparison between the human translation of the original text of one or more languages with purpose to determine exactly and automatically used basic notions, parts of speech and relationships between them, the gender, the number, the tense of the action and tense relationship with other actions.

5. Formalization of a natural language, according to claim 1, characterized also by the step where it generates from unambiguous model of a natural language text a text in an artificial language.

6. A method for determining the basic notions which the humanity uses, necessary for execution of the method given in claim 1, characterized by the following steps:

for each word hi a natural language, a computer finds and extracts its synonyms in a computer dictionary of synonyms;

for each pair of word-synonym a computer compares the descriptions given in a dictionary for the word and for the synonym;

for each two similar texts which contain a given percentage of one and the same words or words-synonyms for a given text, it is supposed that they describe a basic notion;

a computer outputs a list of supposed basic notions and the descriptions which have made that decision;

it is checked in the data base for each supposed basic notion if it is not already registered as it compares discovered in the previous step similar texts to the descriptions of the basic notions in the base and if there is a given percentage of words or words-synonyms it can be considered that the basic notion is already registered and the found description of the basic notion is outputted by the computer and also the other two similar descriptions which are the cause for the search;

an operator checks if the text with outputted by the words coincidence way have a semantic coincidence and if it found such a coincidence he/she decides that the given basic notion is already registered and he/she only adds to the registration one or both words-synonyms which name the basic notion in a certain natural language;

if a given basic notion is not found in the data base, it is added as from the two similar texts is chosen one or the operator specifies the description.

7. A method for addition of a new natural language to the formed base of basic notions, characterized by the following steps:

it is used the method according to the claim 6 for the new language and it is formed second base of basic notions;

from a dictionary from a second language to the first (which already is in the base) are found the possible translations of each name of a basic notion from the new base;

for each translation-word from the first language are extracted the basic notions which can name that word;

it is made pseudo-translations of the description of the basic notion in the second language as it is generated all combinations of substitutions of each word from the description with all possible translations in the first language;

pseudo-translations of the description of the basic notion in the second language are compared in percentage of one and the same words or words-synonyms to the descriptions of the extracted basic notions from the first base;

it is found the best accordance and it is marked;

each found in this way accordance is approved by an operator who decides if found similar descriptions by similar words have semantic accordance;

after approval of the accordance in the second base the basic notion erases, and the list of names of the basic notion from the second language marks that it is in the second language and it adds to the basic notion from the first base;

after processing of all accordances, those basic notions that are still in the second base are registered as new basic notions in the first base or an operator finds their accordance in the first base.

9. Special software according to claim 11, characterized also by the ability to generate explanations in a random level of complexity as using descriptions of the basic notions used in the text, as well as to use recursively the descriptions of the basic notions used for determining of the basic notions in an upper level and to substitute the basic notion with its description.

10. Special software according to claim 11, characterized also by the ability to search in or to process the unambiguous model instead of search in or process the text in the natural language, having in addition the ability to represent the results from the search or processing by a generation of a text in a natural or artificial language or to represent the results as an accordance in the text in a natural language.

11. Special software for implementation of the method according to claim 1, which has the ability to edit text and characterizes with the following abilities:

to can open one connection to the database, where is written previously prepared set of basic notions;

to generate unambiguous models of a text in a natural language using previously prepared basic notions for the given natural language;

to generate from an unambiguous model a text in a natural language;

to be able to set in which natural language to be made the generation from the unambiguous model;

to mark relevant sentences in the original and in the generated texts;

to mark the differences between the relevant sentences in the original and in the generated text;

to represent the description of the basic notion which the computer has chosen for a certain word in a natural language as this representation is made as the words are pointed in the original text or in the text generated according to the unambiguous model;

to be able an operator to change directly or as indicating a synonymous a basic notion which the computer was attached to the word from the text in a natural language;

to be able the operator to indicate the parts of speech and relationship from one part of speech to another;

to be able the operator to indicate the tense relationships between the actions in a complex sentence or the actions in two adjacent sentences;

to be able the operator to indicate what it is substituted by a particular pronoun;

to be able the operator to indicate the external characteristics of the text such as which the subject area of the text is, if it is irony, sarcasm or playing with words.

8. (canceled)