Machine translation: Which tool to use?
To quickly translate Chinese texts (e.g. baidu/wikipedia articles) I use following tools:
- Babylon; rule-based translation engine
- Google translate; statistical engine
- Systran.
In the following an example of the translation quality of each of the tools.
Texttype: popular science article
Subject field: philosophy
Source: http://baike.baidu.com/view/50208.htm, 22/11/2009
性格(character)是一种与社会相关最密切的人格特征,在性格中包含有许多社会道德含义。性格表现了人们对现实和周围世界的态度,并表现在他的行为举止中。
Here the translation of each of the character:
与 yŭ give, take part
社会 shèhuì society
相关 xiāngguān be related
密切 mìqiè close
人格 特征 réngé tèzhēng characteristic, traits
性格 xìnggé character
包含 bāohán contain, include
有许多 yǒu xŭduō have a lot of
道德 dàodé moral
含义 hányì implication
性格(character)是一种与社会相关最密切的人格特征,在性格中包含有许多社会道德含义。
<<Babylon translation>>: Character (character) is a society most close personal character, personality, contains many social ethical implications.
– syntax error: didn’t recognized the attributive construction (是一种与社会相关最密切的人格特征) and made „society“ instead of „traits“ as object of the sentence;
– inconsistency: in the second part of the sentence the program uses „personality“ instead of „character“;
– the last part of the sentence is correct; here the program improved the text choosing the word „ethical“ instead of the more literal translation „social and moral“ (text improvement). Also in the first part of the sentence the program correctly decided not to translate 一种 (a kind), which is a measure word in Chinese and not necessary in the English translation.
性格(character)是一种与社会相关最密切的人格特征,在性格中包含有许多社会道德含义。
The character is the (human) trait which is most related to society and has many ethical implications.
<<Google translate translation>>: Personality (character) is a kind of society most closely related to personality characteristics, personality contains many social and moral implications.
– use of wrong synonym: personality; for personality Chinese has a proper word: 个性 gèxìng;
– uses 一种 which is not necessary (redundant); sea comment above.
– syntax error: didn’t recognized the attributive construction (是一种与社会相关最密切的人格特征) and made „society“ the object of this sentence;
– the second part of the sentence is correct.
性格(character)是一种与社会相关最密切的人格特征,在性格中包含有许多社会道德含义。
<<Systran translation>>: Disposition (character) is one kind with the social related closest personality characteristic, contains in the disposition has many social morals meanings.
– wrong translation: disposition is not character, it is only a part of it;
– the program translates word per word; it also translated 一种 which is in the TT (target text) not necessary.
– source text improvement: in the second part of the sentence the program eliminates the subject (character); good choice;
– I find this translation even more confusing that the translations provided from the other translation machines.
性格表现了人们对现实和周围世界的 态度,并表现在他的行为举止中。
The character performs people’s disposition towards the reality and surroundings through actions and behaviours.
xìnggé biǎoxiàn le rén mén duì xiànshí hé zhōuwéi shìjiè de tàidù, bìng biǎoxiàn zài tā de xíngwéi jŭzhǐ zhōng.
表现 biǎoxiàn perform
对…态度 tàidù attitude
现实 xiànshí reality
周围世界 zhōuwéi shìjiè (surroundings, world)
并 bìng and
行为 xíngwéi 举止 jŭzhǐ behaviour
<<Babylon translation>>: Character performance of the reality and around the world‘ s attitude, and in his behaviour.
– semantic error: the software didn’t recognise the verb (it is perform and not performance);
– missing translation: it didn’t translate 人们, the object of this phrase; also 现实 xiànshí reality has been omitted;
– syntax error: it didn’t recognise the attributive construction (人们的态度) and translated world’s attitude instead of men’s attitude;
<<Google translate translation>>: The performance of the real character of the people in and around the world, attitude and performance in his behavior.
– error: the software didn’t recognise the verb (performance instead of perform);
– to be completed
<<Systran translation>>: The disposition has displayed the people to the reality and surrounding world manner, and displays in his behavior manner.
– to be completed
Here an intro in the problematic of machine translation for the Chinese language.