Recursion, and recursion-like design patterns, are used in the entry schemas of dictionaries to model subsenses and subentries. Recursion occurs when elements of a given type, such as sense, are allowed to contain elements of the same or similar type, such as sense or subsense. This article argues that recursion unnecessarily increases the computational complexity of entries, making dictionaries less easily processable by machines. The article will show how entry schemas can be simplified by re-engineering subsenses and subentries as relations (like in a relational database) such that we only have flat lists of senses and entries, while the is-subsense-of and is-subentry-of relations are encoded using pairs of unique identifiers. This design pattern losslessly records the same information as recursion (including – importantly – the listing order of items inside an entry) but decreases the complexity of the entry structure and makes dictionary entries more easily machine-processable.

中文翻译：

---

避免递归表示字典中的子义和子条目

递归和类似递归的设计模式被用在字典的条目模式中，以对子意义和子条目进行建模。当允许给定类型的元素（例如 sense）包含相同或相似类型的元素（例如 sense 或 subsense）时，就会发生递归。本文认为递归会不必要地增加条目的计算复杂性，使机器更难处理字典。本文将展示如何通过将子意义和子条目重新设计为关系（如在关系数据库中）来简化条目模式，这样我们只有意义和条目的平面列表，而 is-subsense-of 和 is-subentry-关系使用成对的唯一标识符进行编码。