基于句法依赖关系模板的术语相似度计算方法

A Term Similarity Algorithm Based on Context Dependency Relation Pattern

针对现有基于语境特征的术语相似度算法在语境模板生成和匹配过程中存在的不足,提出基于术语的句法依赖关系自动构造术语语境模板,进而通过语境模板匹配计算术语相似度的方法。该方法既能减少语境模板的生成和匹配困难,又将术语语境特征较好地保留在模板中。针对新方法提出具体的实现步骤,并选取基因工程领域实验数据对新方法和现有典型方法进行对比评测。实验证明,新方法在计算效果方面具有明显提升。

Based on the problems in typical term context similarity algorithm, the paper puts forward a new term similarity algorithm which constructs context patterns automatically by sentences dependencies analysis and then computes term similarity by mapping context patterns. The algorithm provides a better way to construet term context patterns. Meanwhile, term context characters are kept well in patterns. The paper also presents the specific implementation steps of new algorithm, and evaluates the algorithm on basis of gene engineering field experiment data set. Experiment result demonstrates that the algorithm has an obvious improvement in computing performanee.