中文摘要：

在协同翻译过程中，辅助译文的质量是影响协同翻译效率的重要因素，而现有辅助译文生成方法并没有考虑用户对辅助译文的个性化需求。由此，提出了通过建立用户模型来提高辅助译文质量的研究思路，根据翻译知识库与用户知识库的相对熵的大小来决策为用户提供哪个模板。实验表明：在使用用户模型后，协同翻译的效率有了明显的提高。

英文摘要：

This new translation mode is called collaborative translation. In the process, a major factor af- fecting the efficiency of collaborative translation is the quality of the aiding translation. However, the sys- tem does not take users＇individual requirements into consideration. Therefore, this paper proposed a system to improve the quality of aiding translation by building user models. The system will choose the template for users according to the relative entropy of the translation knowledge base and the user knowledge base. The experiments show that the user model greatly improves the efficiency of collaborative translation.