Translation complexities

﻿
• Edit text type translation complexities - opens the Text type translation complexities list for editing.
• Text type translation complexities - values of coefficients that define complexity of performing various types of work, depending on the text type and the project type.

• Text type - for varioust types of texts the amount of work that is required to translate them is different: for example, modules contain many similar words and their translation is automated, thus comparing with the literature text the translation of modules is easier.

• Average - an average complexity of translation entire the text, being multiplicated on the length of the text it is used for preliminary average planning in the Translation planning report.

• Dictionary translation - a ratio of time for translation of dictionary entries to time for translation of the literature text of the same length.

• Dictionary proofreading - a ratio of time for proofreading of dictionary entries to time for translation of the literature text of the same length.

• Proofreading whole - a ratio of time for proofreading of entire the text and time to translation of the literature text of the same length.

• Proofreading changes - a ratio of time for proofreading of changes in the text to time for translation of the literature text of the same length.

• Picture - a ratio of time for inserting the translated picture into the text stored files to time for translation of one character of literature text.

• File - a ratio of time for inserting the translated file into the text stored files to time for translation of one character of literature text.

• Comment - a ratio of time for adding a comment describing an error in dictionary entry or text to time for translation of one character of literature text.

• Change verif. - a ratio of average time for processing (replacement or declining) of the translation change suggestion to time for translation of one character of literature text.

To calculate the amount of work to translate texts the following formula is used:

T = DL × CTD + DL × CPD + Min(TL × CPC, FTL × CPF) + PQ × CP + FQ × CF

Where:

T – the amount of translation work, normalized characters.
DL – the length of dictionary entries, characters.
CTD – the complexity coefficient of the dictionary translation for the corresponding text type.
CPD – the complexity coefficient of the dictionary proofreading for the corresponding text type.
TL – the length of changes in the text.
FTL – the full length of the text.
CPC – one of complexity coefficients of proofreading changes in the text for the corresponding text type and translation method. The way how it is calculated is described below.
CPF – one of complexity coefficients of proofreading entire the text for the corresponding text type and translation method. The way how it is calculated is described below.
PQ – the number of pictures that are in stored files of the translated text.
CP – the complexity coefficient of picture translation for the corresponding text type.
FQ – the number of files that are in stored files of the translated text.
CF – the complexity coefficient of file translation for the corresponding text type.

To calculate the amount of work to proofread texts the following formula is used:

P = DL × CPD + Min(TL × CPC, FLT × CPF) + CQ × CC + SQ × CS

Where:

P – the amount of proofreading work, normalized characters.
DL – the length of dictionary entries, characters.
CPD – the complexity coefficient of the dictionary proofreading for the corresponding text type.
TL – the length of changes in the text.
FTL – the full length of the text.
CPC – one of complexity coefficients of proofreading changes in the text for the corresponding text type and translation method. The way how it is calculated is described below.
CPF – one of complexity coefficients of proofreading entire the text for the corresponding text type and translation method. The way how it is calculated is described below.
CQ – the number of comments added for text and dictionary entries of the text.
CC – the complexity coefficient of adding a comment for the corresponding text type.
SQ – the number of translation change suggestions, accepted with replacement or declined with checking that incorrect translation not used in texts.
CS – the complexity coefficient of processing translation change suggestions for the corresponding text type.

The complexity coefficient of proofreading changes in text (CPC) depends on the text translation mode.

It is equal to the complexity coefficient of the text interface proofreading when translating changes in the text if the text is translated in the interface translation project.
It is equal to the complexity coefficient of the text interface proofreading when translating entire the text if the text is translated in the full translation project.

The complexity coefficient of proofreading entire the text (CPF) depends on the text translation mode.

It is equal to the complexity coefficient of the proofreading whole text when translating changes in the text if the text is translated in the interface translation project.
It is equal to the complexity coefficient of the proofreading whole text when translating entire the text if the text is translated in the full translation project.

Below listed several examples of calculation of work amount using both algorithms, supported by the application, and differences in results are described.

For the Metadata object text type:

Total text length: 13,284 characters;
Length of changes in the text: 451 character;
Total length of dictionary entries corresponding to this text: 366 characters;
Average complexity coefficient of translation: 0.15;
The complexity coefficient of dictionary translation (often contain words combined of existing words, but require finding out the meaning): 1.117;
Complecity coefficient of changes proofreading: 0.065.

The amount of work:

The algorithm using average coefficients (only average complexity) 451 * 0.15 = 68 normalized characters.
The algorithm using grainy coefficients by types of work (many words in the dictionary + proofreading changes) 366 * 1.117 + 451 * 0.065 = 438 normalized characters. More because there are few translated dictionary entries in the dictionary.

For the Module text type:

Total text length: 14,896 characters;
Length of changes in the text: 11,476 characters;
Total length of dictionary entries corresponding to this text: 2,364 characters;
Average complexity coefficient of translation: 0.55;
The complexity coefficient of dictionary translation (contains less words combined of existing words than in case of metadata object): 1.631;
Complecity coefficient of changes proofreading: 0.214;
Complecity coefficient of proofreading entire the text: 0.107;
Proofread entire the text if the ratio changes length to the full length is larger than: 0.5.

The amount of work:

The algorithm using average coefficients (only average complexity) 14896 * 0.55 = 8193 normalized characters.
The algorithm using grainy coefficients by types of work (many words in the dictionary + proofreading entire the text as more than 50% of changes) 2364 * 1.631 + 14896 * 0.107 = 5450 normalized characters. Less, because there are many translated dictionary entries in the dictionary.

For the HTML text type:

Total text length: 3,211 characters;
Length of changes in the text: 702 characters;
Number of pictures: 8;
Total length of dictionary entries corresponding to this text: 1,991 character;
Average complexity coefficient of translation: 0.7;
The complexity coefficient of dictionary translation (long phrases that require finding out meaning only once per each dictionary entry): 0.824;
Complecity coefficient of changes proofreading: 0.272;
The complexity coefficient of picture translation: 240 normalized characters.

The amount of work:

The algorithm using average coefficients (only average complexity) 702 * 0.7 = 491 normalized character.
The algorithm using grainy coefficients by types of work (has pictures) 1991 * 0.824 + 702 * 0.272 + 8 * 240 = 3752 normalized characters. More because many changes in the dictionary and has pictures.

﻿