Browser Myanmar fontMyanmar Font Browser
The Zawgyi to Unicode is converted using the Clidi Zawgyi to Unicode translation rule (see "Zawgyi to Unicode Conversion" below). While this is not an offical Google service, we trust you will find Myanmar Tools useful to better serve Myanmar's world. About Myanmar Tools? The Myanmar Tools uses a mechanized study tool to provide very precise results in the detection of Zawgyi versus Unicode.
If a detector uses hand-coded recognition policies, it is vulnerable to label contents in other programming language such as Shan and Mon as Zawgyi if they are actually Unicode. The Myanmar Tools and Zawgyi converting policies are used by Google, Facebook and others to give Myanmar people a great experience. Converting to Unicode, etc.
If you use the getZawgyiProbability function, it calculates a number between 0 and 1 to indicate the probabilities that a given Zawgyi is either Zawgyi or Unicode. If there are long enough sequences, the search engine should give a number very near 0 or 1, but for sequences with only a few digits, the number may be nearer the center.
In this context, use the following helper statistics to adjust THRESHOLD: If Zawgyi is poor (e.g. if a person is allowed to analyze the result), use a low limit like 0.05. It is this barrier that ensures that less than 1% of Zawgyi chords remain undiscovered. When Zawgyi's over-prediction is poor (e.g. if the translation is automatic), you should place a high limit such as 0.95.
If the Unicode string is less than 1%, this ensures that less than 1% of the Unicode string is incorrectly marked. Also note that you want to adjust your trigger values to the string distributions in your entry datas. If your entry information is prejudiced towards Unicode, for example, to decrease falsification positive, you can specify a higher Zawgyi limit than if your entry information is prejudiced towards Zawgyi.
Finally, the best way to select threshold values is to obtain a series of labelled strands that are indicative of the information that the sensor will process, calculate their values and adjust the threshold values to your required precision-recall relationship. The same Zawgyi probabilities are obtained when a given expression contains a non-Burmese fix as if the fix had been deleted.
A number of character chains are the same in V and A; this can occur if the character chain usually contains a few diacritical vocals. Any value for such sequences can be returned by the detectors. In this case, the operator can run the sequence through a convertor and see if the exit of the convertor is the same as the entry of the transform.
Zawgyi detector's working example was used to train several Megabyte of website information over the web. Notice that you must at least make sure that you have received Unicode and Zawgyi files; the folder should contain my. trxt, my-t-t-d0-zawgyi. trxt and shn.txt. It is often useful to change from one code to another once you have discovered that a text is Unicode or Zawgyi.
Even though this is not a custom translation supported repository, the Common Locale Data Repository (CLDR) releases a Zawgyi to Unicode transliter in the format of translation snippets. A lot of other programming language, such as Python, Ruby and PHP, have wrappers via ICU4C, which means that you can also use the Zawgyi convertor in these language.