https://journals.colorado.edu/index.php/lilt/issue/feed 2022-07-18T14:41:07-07:00 Martha Palmer martha.palmer@colorado.edu Open Journal Systems <p><em>Linguistic Issues in Language Technology</em> (<em>LiLT</em>) is an open-access, open-data journal that focuses on relationships between linguistic insights, which can prove valuable to language technology, and language technology, which can enrich linguistic research. The Editorial Board of <em>LiLT</em> believes that, in conjunction with machine learning and statistical techniques, deeper and more sophisticated models of language and speech are needed to make significant progress in newly emerging areas of computational language analysis. <em>LiLT</em> provides a forum for such work and takes an eclectic view on methodology.</p> https://journals.colorado.edu/index.php/lilt/article/view/939 2021-12-01T03:05:36-07:00 Jennifer Sikos jennifercarlso@gmail.com Michael Roth michael.roth@ims.uni-stuttgart.de Sebastian Padó pado@ims.uni-stuttgart.de

<p>A recent research direction in computational linguistics involves efforts to make the field, which used to focus primarily on English, more multilingual and inclusive. However, resource creation often remains a bottleneck for many languages, in particular at the semantic level. In this article, we consider the case of frame-semantic annotation. We investigate how to perform frame selection for annotation in a target language by taking advantage of existing annotations in different, supplementary languages, with the goal of reducing the required annotation effort in the target language. We measure success by training and testing frame identification models for the target language. We base our selection methods on measuring frame transferability in the supplementary language, where we estimate which frames will transfer poorly, and therefore should receive more annotation, in the target language. We apply our approach to English, German, and French – three languages which have annotations that are similar in size as well as frames with overlapping lexicographic definitions. We find that transferability is indeed a useful indicator and supports a setup where a limited amount of target language data is sufficient to train frame identification systems.</p>

2022-07-18T00:00:00-07:00 Copyright (c) 2022 Jennifer Sikos, Michael Roth, Sebastian Padó