Heitmeier Maria, Chuang Yu-Ying, Baayen R Harald
Department of Linguistics, Eberhard-Karls Universität, Tübingen, Germany.
Front Psychol. 2021 Nov 15;12:720713. doi: 10.3389/fpsyg.2021.720713. eCollection 2021.
This study addresses a series of methodological questions that arise when modeling inflectional morphology with Linear Discriminative Learning. Taking the semi-productive German noun system as example, we illustrate how decisions made about the representation of form and meaning influence model performance. We clarify that for modeling frequency effects in learning, it is essential to make use of incremental learning rather than the end-state of learning. We also discuss how the model can be set up to approximate the learning of inflected words in context. In addition, we illustrate how in this approach the wug task can be modeled. The model provides an excellent memory for known words, but appropriately shows more limited performance for unseen data, in line with the semi-productivity of German noun inflection and generalization performance of native German speakers.
本研究探讨了使用线性判别学习对屈折形态进行建模时出现的一系列方法学问题。以半生成性的德语名词系统为例，我们说明了关于形式和意义表示所做的决策如何影响模型性能。我们阐明，为了对学习中的频率效应进行建模，利用增量学习而非学习的最终状态至关重要。我们还讨论了如何设置模型以近似在语境中对屈折词的学习。此外，我们说明了在这种方法中如何对“wug任务”进行建模。该模型对已知单词有出色的记忆，但对于未见过的数据表现出更有限的性能，这与德语名词屈折的半生成性以及以德语为母语者的泛化性能相符。
