This paper describes the ongoing process of creating a computational morphological model of Plains Cree, a language native to North America, making use of finite-state machines, and with a focus on verbs. We cover prior linguistic theoretical and descriptive models of Plains Cree, moving on to the computational implementation of (chiefly) inflectional phenomena, followed by relevant morphophonological processes. We evaluate the performance of our computational implementation with a hand-verified corpus of Plains Cree, and present a discussion of the morphological complexity found in the corpus, as compared to that of our model and its theoretical underpinnings. The results of this evaluation and research into natural language use inform us about the practical extent of morphological complexity for a polysynthetic language, and allow us to identify avenues for improvement of the model. Finally, this computational model for Plains Cree offers the opportunity to create various digital tools and applications for language users for the maintenance and revitalization of this language in the 21st century.

中文翻译：

---

从Plains Cree动词的计算模型中学习

本文介绍了使用有限状态机并着重于动词来创建北美洲本地语言Plains Cree的计算形态模型的过程。我们讨论了Plains Cree的先前语言理论和描述模型，然后继续（主要）拐弯现象的计算实现，然后介绍相关的形态学过程。我们用一个手工验证的Plains Cree语料库评估了我们的计算实现的性能，并与该模型及其理论基础进行了比较，讨论了在该语料库中发现的形态复杂性。评估结果和对自然语言使用的研究使我们了解了一种多合成语言的形态复杂性的实际范围，并允许我们确定改进模型的途径。最终，Plains Cree的这种计算模型为有机地为语言用户创建各种数字工具和应用程序提供了机会，以在21世纪维护和振兴这种语言。