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

Social media such as Twitter provide valuable information to crisis managersand affected people during natural disasters. Machine learning can helpstructure and extract information from the large volume of messages sharedduring a crisis; however, the constantly evolving nature of crises makeseffective domain adaptation essential. Supervised classification is limited byunchangeable class labels that may not be relevant to new events, andunsupervised topic modelling by insufficient prior knowledge. In this paper, webridge the gap between the two and show that BERT embeddings finetuned oncrisis-related tweet classification can effectively be used to adapt to a newcrisis, discovering novel topics while preserving relevant classes fromsupervised training, and leveraging bidirectional self-attention to extracttopic keywords. We create a dataset of tweets from a snowstorm to evaluate ourmethod's transferability to new crises, and find that it outperformstraditional topic models in both automatic, and human evaluations grounded inthe needs of crisis managers. More broadly, our method can be used for textualdomain adaptation where the latent classes are unknown but overlap with knownclasses from other domains.

Bridging the gap between supervised classification and unsupervised topic modelling for social-media assisted crisis management