Using BERT Embeddings to Model Word Importance in Conversational Transcripts for Deaf and Hard of Hearing Users
Abstract
Deaf and hard of hearing individuals regularly rely on captioning while watching live TV. Live TV captioning is evaluated by regulatory agencies using various caption evaluation metrics. However, caption evaluation metrics are often not informed by preferences of DHH users or how meaningful the captions are. There is a need to construct caption evaluation metrics that take the relative importance of words in transcript into account. We conducted correlation analysis between two types of word embeddings and human-annotated labelled word-importance scores in existing corpus. We found that normalized contextualized word embeddings generated using BERT correlated better with manually annotated importance scores than word2vec-based word embeddings. We make available a pairing of word embeddings and their human-annotated importance scores. We also provide proof-of-concept utility by training word importance models, achieving an F1-score of 0.57 in the 6-class word importance classification task.
Akhter Al Amin
Software Engineer at Amazon
Saad Hassan
Assistant Professor
My research interests include human-computer interaction (HCI), accessibility, and computational social science.
Cecilia 0. Alm
Professor at RIT
