Leveraging WordNet Paths for Neural Hypernym Prediction

Proceedings of the 28th International Conference on Computational Linguistics, pages 3007–3018 Barcelona, Spain (Online), December 8-13, 2020 3007 Leveraging WordNet Paths for Neural Hypernym Prediction Yejin Cho∗,1 Juan Diego Rodriguez∗,2 Yifan Gao3 Katrin Erk1 1Department of Linguistics 2Applied Research Laboratories The University of Texas at Austin3Department of Mathematics {ycho, yifan233, katrin.erk}@utexas.edu juan.rodriguez@arlut.utexas.edu Abstract We formulate the problem of hypernym prediction as a sequence generation task, where the se- quences are taxonomy paths in WordNet. Our experiments with encoder-decoder models show that training to generate taxonomy paths can improve the performance of direct hypernym pre- diction. As a simple but powerful model, the hypo2path model achieves state-of-the-art perfor- mance, outperforming the best benchmark by 4.11 points in hit-at-one (H@1). 1 Introduction Hypernymy, or the IS-A relation, is one of the most important lexical relations. It is used to create tax- onomies of terms and it is the main organizational criterion of nouns and verbs in WordNet (Fellbaum, 1998). Learning hypernymy is also important in practice, as knowing a word’s hypernyms gives an ap- proximation of its meaning, and enables inferences in downstream tasks such as question answering and reading comprehension. Predicting hypernymy is still a challenging task for word embeddings (Pinter and Eisenstein, 2018; Bernier-Colborne and Barriere, 2018; Nickel and Kiela, 2018) and previous stud- ies have shown that it is more difficult to predict hypernymy than other lexical relations (Balaˇzevi´c et al., 2019; Allen et al., 2019). Hypernymy prediction is often evaluated against a given taxonomy, typically WordNet (Fellbaum, 1998). The main hypothesis that we pursue in this paper is that knowledge of this taxonomy, in particular of taxonomy paths, will be helpful for hypernymy prediction. So we introduce two simple encoder- decoder based models for hypernym prediction that make use of information in the full taxonomy paths. There has been much recent work on modeling lexical relations based on distributed representations (Pinter and Eisenstein, 2018; Bernier-Colborne and Barriere, 2018; Nickel and Kiela, 2018). However, the task formulations and evaluation datasets have differed widely, making it hard to compare different approaches. We focus on evaluating on hypernymy, rather than jointly on many relations, which can mask strong performance differences across relations. We evaluate our encoder-decoder models against several previous models that have not been evaluated in the same setting before. Like many other approaches, we use WordNet as the basis for our experiments. We formulate the task as the task of finding the correct point to attach a new node (synset) to the WordNet taxonomy. We build on the existing WN18RR dataset (Dettmers et al., 2018), but filter its hypernymy pairs to produce WN18RR-hp, a subset that is leak-free with respect to approaches that use taxonomy paths during training, as we do. We find that one of our new models, hyper2path, achieves state-of-the-art performance on hypernym prediction on WN18RR-hp, exceeding the best performance of benchmark models by 4.11 points in accuracy of the highest-ranked prediction (hit-at-one, H@1). In particular, we observe the greatest per- formance gain in noun hypernym prediction, …

Content truncated to 500 words