本文首先分析了德语中的语义歧义现象，然后提出了几种借助配价和语义信息进行消歧的策略。这些策略目前都已应用于同济大学开发的ＴＪＴＩＴＲ德汉机器翻译系统中。实践证明，它们不仅较好地解决了机器翻译中的语义歧义问题，而且大大提高了系统运行的效率 This paper first analyzes the semantic ambiguity in German and then proposes several strategies for disambiguation based on valence and semantic information. These strategies are currently used in TJTITR German-Chinese machine translation system developed by Tongji University. Practice has proved that they not only better solve the problem of semantic ambiguity in machine translation, but also greatly improve the efficiency of system operation