A Proposal for Linguistic Similarity Datasets Based on Commonality Lists
This work addresses a methodological issue in computational linguistics for researchers developing similarity datasets, though it appears incremental as it builds on existing psychological insights.
The authors tackled the problem of ambiguous similarity scores in multi-category linguistic datasets by proposing a new collection procedure where humans list commonalities and differences instead of assigning numerical scores, aiming to improve the evaluation of meaning representation models.
Similarity is a core notion that is used in psychology and two branches of linguistics: theoretical and computational. The similarity datasets that come from the two fields differ in design: psychological datasets are focused around a certain topic such as fruit names, while linguistic datasets contain words from various categories. The later makes humans assign low similarity scores to the words that have nothing in common and to the words that have contrast in meaning, making similarity scores ambiguous. In this work we discuss the similarity collection procedure for a multi-category dataset that avoids score ambiguity and suggest changes to the evaluation procedure to reflect the insights of psychological literature for word, phrase and sentence similarity. We suggest to ask humans to provide a list of commonalities and differences instead of numerical similarity scores and employ the structure of human judgements beyond pairwise similarity for model evaluation. We believe that the proposed approach will give rise to datasets that test meaning representation models more thoroughly with respect to the human treatment of similarity.