Determining Differences of Granularity between Cross-Dictionary Linked Senses

We present a machine learning model classifying the relation between two linked dictionary senses based on the senses’ granularity. The relations are perfect, narrower/wider, and partial. The model yields an overall accuracy of 86% and significantly outperforms a rule-based algorithm serving as the baseline.We present a machine learning model classifying the relation between two linked dictionary senses based on the senses’ granularity. The relations are perfect, narrower/wider, and partial. The model yields an overall accuracy of 86% and significantly outperforms a rule-based algorithm serving as the baseline.