Abstract: Description words in the text writing process are the core fields of teaching knowledge in automatic writing’s teaching system. According to the characteristics that multilingual sources lead to multiple data types of knowledge points, complex computing structure, and the large amount of knowledge, the method about automatic generation of description words in the text writing process is proposed. Description words are composed of the leading stroke order and constraint stroke order. Coding space relationships of the various elements are given. The algorithm of automatic generation of description words is designed and implemented. The experimental results on the fusion of Chinese, English, and pinyin show that descriptive word structure can satisfy the writing process expression with different language and the generation method can accurately identify various elements of knowledge. Compared with the manual method, it not only enhances efficiency and accuracy of knowledge, but also effectively reduces the degree of redundancy.

Key words: character writing process, description words, automatic generation, multilingual integration, writing teaching, Elements

摘要： 文字书写过程描述字是文字书写自动教学系统教学知识点的核心字段。针对多语源导致知识点数据类型多、计算结构复杂及知识量大等特点，提出文字书写过程描述字自动生成方法。描述字由主导笔顺和约束笔顺组成；给出各类要素元编码空间关系；设计与实现描述字自动生成算法。汉字、英文、汉语拼音等文种文字融合实验及其分析表明，描述字结构能满足不同文种文字书写过程表达，生成方法能准确识别各种知识要素，与手工法相比不但提高了工作效率、知识的准确率，而且有效降低了信息冗余度。

关键词: 文字书写过程, 描述字, 自动生成, 多文种融合, 文字书写教学, 要素元