English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Deaf and hard of hearing individuals regularly rely on captioning whilewatching live TV. Live TV captioning is evaluated by regulatory agencies usingvarious caption evaluation metrics. However, caption evaluation metrics areoften not informed by preferences of DHH users or how meaningful the captionsare. There is a need to construct caption evaluation metrics that take therelative importance of words in a transcript into account. We conductedcorrelation analysis between two types of word embeddings and human-annotatedlabeled word-importance scores in existing corpus. We found that normalizedcontextualized word embeddings generated using BERT correlated better withmanually annotated importance scores than word2vec-based word embeddings. Wemake available a pairing of word embeddings and their human-annotatedimportance scores. We also provide proof-of-concept utility by training wordimportance models, achieving an F1-score of 0.57 in the 6-class word importanceclassification task.

Using BERT Embeddings to Model Word Importance in Conversational Transcripts for Deaf and Hard of Hearing Users