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/

Zendy Plus

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

Zendy Tools

Zendy Open

The human ability of deep cognitive skills are crucial for the development ofvarious real-world applications that process diverse and abundant usergenerated input. While recent progress of deep learning and natural languageprocessing have enabled learning system to reach human performance on somebenchmarks requiring shallow semantics, such human ability still remainschallenging for even modern contextual embedding models, as pointed out by manyrecent studies. Existing machine comprehension datasets assume sentence-levelinput, lack of casual or motivational inferences, or could be answered withquestion-answer bias. Here, we present a challenging novel task, tropedetection on films, in an effort to create a situation and behaviorunderstanding for machines. Tropes are storytelling devices that are frequentlyused as ingredients in recipes for creative works. Comparing to existing movietag prediction tasks, tropes are more sophisticated as they can vary widely,from a moral concept to a series of circumstances, and embedded withmotivations and cause-and-effects. We introduce a new dataset, Tropes in MovieSynopses (TiMoS), with 5623 movie synopses and 95 different tropes collectingfrom a Wikipedia-style database, TVTropes. We present a multi-streamcomprehension network (MulCom) leveraging multi-level attention of words,sentences, and role relations. Experimental result demonstrates that modernmodels including BERT contextual embedding, movie tag prediction systems, andrelational networks, perform at most 37% of human performance (23.97/64.87) interms of F1 score. Our MulCom outperforms all modern baselines, by 1.5 to 5.0F1 score and 1.5 to 3.0 mean of average precision (mAP) score. We also providea detailed analysis and human evaluation to pave ways for future research.

Situation and Behavior Understanding by Trope Detection on Films