Finding IndividualWord Sense Changes and their Delay in Appearance

Abstract

We present a method for detecting word sense changes by utilizing automatically induced word senses. Our method works on the level of individual senses and allows a word to have e.g. one stable sense and then add a novel sense that later experiences change. Senses are grouped based on polysemy to find linguistic concepts and we can find broadening and narrowing as well as novel (polysemous and homonymic) senses. We evaluate on a testset, present recall and estimates of the time between expected and found change.

Effects of Pre- and Post-Processing on type-based Embeddings in Lexical Semantic Change Detection

“…Hence, we recommend to obtain a small annotated sample of target words for the target corpora and to tune pre-training, model and post-processing parameters on the sample before performing predictions for semantic changes on unseen data. With the recent upsurge of digitized historical corpora and diachronic semantic annotation efforts (Tahmasebi and Risse, 2017;Schlechtweg et al, 2018Basile et al, 2020; this may often be a likely and feasible scenario.…”

Lexical Semantic Change Discovery

“…As mentioned above, previous work mostly focuses on creating data sets or developing, evaluating and analyzing models. A common approach for evaluation is to annotate target words selected from dictionaries in specific corpora (Tahmasebi and Risse, 2017;Schlechtweg et al, 2018;Perrone et al, 2019;Basile et al, 2020;Rodina and Kutuzov, 2020;. Contrary to this, our goal is to find 'undiscovered' changing words and validate the predictions of our models by human annotators.…”

Diachronic Sense Modeling with Deep Contextualized Word Embeddings: An Ecological View

“…Tang et al (2016) attempted to cluster the contexts to find senses, and to classify the senses into different change types. Tahmasebi and Risse (2017) exploit curvature clustering algorithm to induce word senses and track the change of them.…”