CafeTran Help

Java memory size (MB)

When CafeTran is launched, Java allocates a computer memory for all its operations. The maximum size of this memory in megabytes can be set in this field. It should be large enough to load the resources used in the translation process, but it cannot be larger than the total physical memory (RAM) available in the computer. Moreover, the 32-bit Java version has the maximum memory limit of around 2 gigabytes. Therefore, it is recommended to use 64-bit Java on the 64-bit system architecture when loading huge translation memories or glossaries into CafeTran. The default value is 1024 megabytes and it should be increased if the program hangs when loading memories or glossaries.

Prefix matching (%)

This option determines the default settings of the Prefix matching function. When the box is checked, Translation Memories will have the Prefix matching turned on at the start of loading. From the drop-down box select the desired percentage of the word's length as its prefix or the fixed prefix length which is the minimal prefix length of the word.

Minimal prefix length

The field determines the minimal length of prefixes of the words processed by the Prefix matching function.

Memory matching type

The selection lets you decide the default matching type when the Translation Memory is loaded. The Fuzzy & Subsegment selection makes the program search for fuzzy full segment matches as well as exact and fuzzy subsegment matches. In the Fuzzy selection, the program searches for fuzzy full segment matches and exact subsegment matches, while the fuzzy subsegment matching is turned off. The selection of the Detailed matching should be made when the source language does not have a space as a word separator e.g. Japanese or Chinese. When this option is selected, CafeTran performs the detailed character-based analysis of the source text to find matches in source language segments.

Subsegment to Auto threshold

Looking for a fuzzy match of the source subsegment, CafeTran counts the found fragments in the target segments for the given source subsegment. When the occurrence count reaches the defined number called the ‘Threshold Number’, the program will send this target language fragment to the Auto-assembling function as the translation of the source language subsegment. You can set this Threshold Number in this field.

Subsegment to Virtual threshold

In this field, you may define the occurrence threshold as above. When this occurrence number is reached, CafeTran sets the found target language fragment as the constant translation of the given subsegment, thus avoiding complex text analysis when the subsegment occurs again. This target language fragment is placed in the virtual memory to be retrieved quickly when necessary. This number determines how many hits CT needs for a subsegment guess to be treated as an exact subsegment match. Since this isn’t an ‘exact’ match in the full sense of the word, CT calls it a ‘virtual’ match. Note: You can lower the possibility of CT guessing incorrectly by increasing this number.

Subsegment look-up limit

Set the maximum number of the matches found for the given subsegment. When this number is reached, the program stops searching for any more matches.

Auto-assembling insert threshold (%)

The field lets you define the accuracy of the Auto-assembling translation at which the program inserts it automatically into the target language box.

Fuzzy match insert threshold (%)

The field lets you define the accuracy of the best fuzzy match at which the program inserts it automatically into the target language box.

Function words threshold (%)

CafeTran filters out frequently occurring words which do not need to be translated into the target language as single fragments. In this field you can set the default frequency of occurrence of the words in the source text above which the program treats them as function words and they are omitted by the Auto-assembling function.

Fuzzy match threshold (%)

Sets the minimum accuracy of the fuzzy matches which are displayed by CafeTran.

Fuzzy match display limit:

Sets the maximum number of the fuzzy matches displayed by CafeTran.

Memory integration:

Select the default integration of Translation Memories in the workflow. In the Automatic integration, CafeTran searches for matches every time the current source segment is displayed in the source language box. When the Pretranslation option is selected, the program searches for matches of all source segments once at the start of the translation and displays the result when the current source segment is set in the source language box. In the Manual integration, the user needs to press the Translate button to make the program search memories for matches of the current segment.

Additional space characters (Unicode numbers)

U+00A0,U+2009 Here you can insert additional Unicode characters that should be treated as spaces. This is important for term recognition and hence for auto-assembling.

Do not match

Specify the list of characters that hinder the correct recognition of words. These are usually punctuation characters. CafeTran ignores these characters while searching for matches of the source segments in Translation Memories and Glossaries.

RESET button

Pressing this button resets all CafeTran settings to default values. You need to restart the program to take effect.

OK button

Press this button to accept changes in CafeTran options and close the Options window.

CANCEL button

Press this button to discard any changes and close the Options window.

To insert:

Subsegment minimal length:
CafeTran 'guesses' the meaning of source subsegments only if they are
equal or longer than the set minimal length in characters.

Subsegment minimal length difference:
CafeTran 'guesses' the meaning of a source subsegment only if the length
difference between the source subsegment and 'guessed' target subsegment
is no lower than the set percentage.

Search display limit:
CafeTran limits the display of found segments with a searched word to the
set number to avoid displaying hundreds of superfluous segments that also
contain the word.

Previous Next