Looking for a partner to commercialize of the project, my expectations I will present the persons concerned.

Contact details: e-mail: virp2@poczta.onet.pl


Computer Translator with High Accuracy Ratio of the Text Translation
into another Language or Languages


The Object of the Invention. The object of the invention is a computer translator with high accuracy ratio of the text
translation into another language or languages, which can be widely used as a translating program installed on a
PC, in the next-generation mobile phone, in an electronic translator, as well as on the operator server that offers the
ability to translate text with the translator of the present invention. 

State of the Art. Existing computer translators are largely unreliable as far as quality of translation is concerned.
Some of them, such as Google's translator are unreliable because they are not able to correctly identify ambiguity of
the meanings of words used and then differentiate and select the proper meaning from among them and/or
determine the correct inflected form. Other translators do not have vocabulary large enough to satisfy users,
especially professional ones. Other translators combine all the above deficiencies.
Google offers translating texts into many languages. The translation process may proceed in three ways. In version
one a person who translates the text, does not know the language of the source text he or she intends to translate.
The text is translated into the target language, unknown to the person who translates. In this case, it is possible to
remove the ambiguity of words or phrases only in the translated target text. It is impossible however to verify the
correctness of removal of the ambiguity of words or phrases in the translated target text if the person who translates
does not know the language of the source text. Thus the correct translation of the source text is impossible. In
version two a person who translates the text, knows the language of the source text he or she intends to translate.
The text is translated into the target language, unknown to the person who translates.  In this case, it is possible to
remove the ambiguity of words or phrases only in the translated target text. It is impossible however to verify the
correctness of removal of the ambiguity of words or phrases in the translated target text if the person who translates
does not know the language of the source text. Thus the correct translation of the source text is impossible. In
version three a person who uses the Internet, with intention to translate text on the selected web page enters the
Internet address of this page into the translator window, then selects the command ‘Translate’, which results in
indicated web page translation and then displaying it on the Google subpage. The result is a translation of the target
text with ‘suggest a better translation’ option but without the option to remove the ambiguity of words and phrases.
This version also is burdened with deficiency which is inability to verify the accuracy of the translated text by the
person who translates, without knowing the source language.

The Essence of the Invention. Computer translator with high accuracy ratio of the text translation into another
language or languages allows distinguishing the ambiguous meanings of words used and selecting the proper
meaning from among them. The translator of the present invention is dedicated to people who care about high
quality and accuracy of the text translation into a foreign language, and while working with the text in order to do this
they intend to edit the source text written in a language they know by eliminating the ambiguity of words and phrases.
This innovation is the main advantage of the translator of the present invention, which results in obtaining a true and
correct translation into any selected target language.
The translator of the present invention may have multiple application platforms, but the principle of its operation is
based on a single fundamental mechanism. This mechanism is based on adequate preparation of the source text by
removing the ambiguity of words and phrases, which is then stored as a universal code, called the kernel language,
so we get a true and correct translation of the kernel language into any target language, available in the program
configuration. Obtaining the kernel language is a strong asset of the invention, since the kernel language allows
translation of the ‘purified’ text into multiple languages ​​while maintaining perfect translation, allowing you to obtain
almost 100% effective error-free translation of the source text.
Ambiguity of words and phrases is removed as follows: translating program interface is divided into two windows,
one of which is designed for typing or pasting the source text, while the second window is designed to display text in
the selected target language in real time or after the process of eliminating the ambiguity of words and phrases. The
person who translates may at any time copy the text from the target text window and use it. Words and phrases in
the source text window, which are ambiguous, are highlighted and by indicating that word or phrase, possible
meanings of a word or phrase are displayed. Selection the proper meaning of a word or phrase is carried out by
selecting a synonym or description of the word or phrase. The word ‘jumper’ that can have three meanings may be
an example: a person or animal that jumps, a metal bridge to close a circuit, a kind of sweater. Each word in the
source window can also be marked as a proper name which should not be translated. Once the proper meaning of
the word is selected, translator stores it in the appropriate form of the kernel language. The kernel language may
take the form of the Polish language with tags. In case words are unambiguous, there are tags as follows: ‘1’ tag for
the word, which is subject to translation, and the ‘2’ tag for the same word, which is a proper name and should not
be translated. In case the word is ambiguous, that word is assigned the appropriate number of tags that
corresponds to the number of meanings of this word plus a proper name, which should not be translated. Thus in
case of the word ‘jumper’, in the kernel language there are four possible meanings of the word: ‘jumper1’ - a person
or animal that jumps, ‘jumper2’ - a metal bridge to close a circuit, ‘jumper3’ - a kind of sweater, ‘jumper4’  - a proper
name.
The source text can also contain ambiguous phrases such as idioms, like ‘you scratch my back and I’ll scratch
yours’ or ‘pick holes’. In such a case, the person editing the source text indicates whether this phrase is an idiom or
words have different meaning. The kernel language, defined by removing the ambiguity of words, allows to precisely
translating the text into any language, present in the translator’s ‘library’ of the present invention. Translator’s library
is made up of at least two languages ​​and their quantity depends on the software version and configuration.
Ultimately, translator of the present invention may precisely translate of all world’s natural languages from any
world’s natural language into any world’s natural language. Removal of ambiguity of words and phrases, which is
‘editing’ the kernel language, may proceed from the level of any natural language located in the present invention
translator’s library. Removal of ambiguity of words from the source text and storing it as a kernel language requires
the use of appropriate algorithms that allow creating the kernel language in compliance with all rules of correctness
of the language. The Polish language is the exception, if the words of the Polish language with tags are the kernel
language, since in this case all the rules of the correctness of the language are observed. Due to the heterogeneous
structure of sentences in different natural languages, the text brought to the form of the kernel language, is translated
into the selected target language using individual algorithms for each target language. In order to obtain high
translation efficiency of the text, the kernel language words in addition to the word meaning tags, might have
additional tags that define the following: parts of speech - noun, verb, adjective, etc., parts of sentence - subject,
predicate, etc., declinations, conjugations, or grammatical genders. Enriching the kernel language words with the
above information allows used algorithms to precisely construct sentences of the translated text.  

Sample Embodiment of the Present Invention. The invention is shown in the embodiment where the computer
program was developed. Its interface is divided into two windows, one of which is designed for typing or pasting the
source text, while the other window is designed for displaying text in the selected target language in real time or after
the completion of eliminating ambiguity of words and phrases. The person who translates may at any time, copy the
text from the target window and use it. Ambiguous words and phrases in the source text window are highlighted and
by indicating the word or phrase, possible meanings of a word or phrase are displayed. The correct meaning of a
word or phrase is done by selecting a synonym or description of the word or phrase. The word ‘jumper’ that can
have three meanings may be an example: 1 - ‘a jumper’ meaning a person or animal that jumps, 2 - ‘a jumper’
meaning a metal bridge to close a circuit, 3 ‘a jumper’ meaning a kind of sweater.  Each word in the source window
can also be tagged as a proper name, which should not be translated. Once the proper meaning of the word is
selected, translator stores it in the appropriate form of the kernel language. The kernel language may take the form
of the Polish language with tags.  In case words are unambiguous, there are tags as follows: ‘1’ tag for the word,
which is subject to translation, and the ‘2’ tag for the same word, which is a proper name and should not be
translated. In case the word is ambiguous, that word is assigned the appropriate number of tags that corresponds to
the number of meanings of this word plus a proper name, which should not be translated.  Thus in case of the word
‘jumper’, in the kernel language there are four possible meanings of the word: ‘jumper1’ - a person or animal that
jumps, ‘jumper2’ - a metal bridge to close a circuit, ‘jumper3’ - a kind of sweater, ‘jumper4’  - a proper name.
The source text can also contain ambiguous phrases such as idioms, like ‘you scratch my back and I’ll scratch
yours’ or ‘pick holes’. In such a case, the person editing the source text indicates whether this phrase is an idiom or
words have different meaning.
The kernel language, defined by removing the ambiguity of words, allows to precisely translating the text into any
language, present in the translator’s ‘library’ of the present invention. Translator’s library is made up of at least two
languages ​​and their quantity depends on the software version and configuration. Ultimately, translator of the present
invention may precisely translate of all world’s natural languages from any world’s natural language into any world’s
natural language. Removal of ambiguity of words and phrases, which is ‘editing’ the kernel language, may proceed
from the level of any natural language located in the present invention translator’s library. Removal of ambiguity of
words from the source text and storing it as a kernel language requires the use of appropriate algorithms that allow
creating the kernel language in compliance with all rules of correctness of the language. The Polish language is the
exception, if the words of the Polish language with tags are the kernel language, since in this case all the rules of the
correctness of the language are observed. Due to the heterogeneous structure of sentences in different natural
languages, the text brought to the form of the kernel language, is translated into the selected target language using
individual algorithms for each target language. In order to obtain high translation efficiency of the text, the kernel
language words in addition to the word meaning tags, might have additional tags that define the following: parts of
speech - noun, verb, adjective, etc., parts of sentence - subject, predicate, etc., declinations, conjugations, or
grammatical genders. Enriching the kernel language words with the above information allows used algorithms to
precisely construct sentences of the translated text.

The Use of the Invention. The computer translator with high accuracy ratio of the text translation into another
language or languages may be widely used as a translating program installed on a PC, in the next-generation mobile
phone, in an electronic translator, as well as on the operator server that offers the ability to translate text with the
translator of the present invention. 
The translator of the present invention may also be a computer program properly modifying the HTML, XHTML and
PHP web pages code, allowing an adapted Web browser to automatically translate text into multiple languages. The
translator of the present invention may also be instant messaging software to allow the real-time chatting with very
high true and correct conversation translation. The translator of the present invention may also be an e-mail program
to write correctly translated email letters.
The translator of the present invention may also be part of the software development tools used in a Delphi or
Borland C + +, giving the ability to create computer programs with ‘universal’ language platform, so that the created
program could be universal from the linguistic point of view, because by selecting the relevant language version, it
would be readable in multiple natural languages ​​of the world. The translator of the present invention can also be
used when working on the development of the computer operating system and other electronic equipment, which
ultimately allows, by selecting the relevant language version, this system to be used by people of many nationalities,
each of which uses different natural language ​​of the world. The translator of the present invention may be a
breakthrough tool worldwide, giving all users using the translator opportunity for error-free communication without
borders. 
Greetings

Radosław Pełka
Copyright © 2013 - 2018 by "Radosław Pełka" • All Rights Reserved • cookies policy
Radosław Pełka
Radosław Pełka na LinkedIn