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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

News articles both shape and reflect public opinion across the politicalspectrum. Analyzing them for social bias can thus provide valuable insights,such as prevailing stereotypes in society and the media, which are oftenadopted by NLP models trained on respective data. Recent work has relied onword embedding bias measures, such as WEAT. However, several representationissues of embeddings can harm the measures' accuracy, including low-resourcesettings and token frequency differences. In this work, we study what kind ofembedding algorithm serves best to accurately measure types of social biasknown to exist in US online news articles. To cover the whole spectrum ofpolitical bias in the US, we collect 500k articles and review psychologyliterature with respect to expected social bias. We then quantify social biasusing WEAT along with embedding algorithms that account for the aforementionedissues. We compare how models trained with the algorithms on news articlesrepresent the expected social bias. Our results suggest that the standard wayto quantify bias does not align well with knowledge from psychology. While theproposed algorithms reduce the~gap, they still do not fully match theliterature.

No Word Embedding Model Is Perfect: Evaluating the Representation Accuracy for Social Bias in the Media