Finding non-recorded senses is important for dictionary maintenance, and using automatic methods helps reduce manual efforts. In this study, we use automatic Word Sense Induction (WSI) to compare recorded sense numbers among a sample of headwords in a comprehensive Swedish monolingual dictionary with induced sense numbers for the same words in a Swedish corpus.

We propose this as a simple technique to identify words that should be prioritized for post-hoc manual checking, which can be done through an easy-to-use online interface that does not require programming knowledge.

A thorough manual evaluation of the proposed methodology demonstrates—statistically—that using automatic WSI increases the odds of finding non-recorded senses compared to a random selection of words. We further (i) evaluate predictions according to their potential inclusion in the dictionary, providing strong evidence of the method’s usefulness in practical lexicography, and (ii) analyze model predictions in depth to highlight directions for future improvement.

Finally, we integrate lessons learned from our analysis into a large-scale prediction effort, providing the first high-quality, large-scale WSI predictions for Swedish. These predictions represent a valuable resource for future research in Swedish lexicography.