Matches in SemOpenAlex for { <https://semopenalex.org/work/W4328006272> ?p ?o ?g. }
Showing items 1 to 66 of
66
with 100 items per page.
- W4328006272 abstract "Clickbait has been a nerve-wracking phenomenon because of its enticing, deceitful features driven by commercial purposes after Internet has been widely used. When machine learning methods begin to thrive in Natural Language Processing(NLP) area, clickbait detection task based on machine learning methods has been studied. And later, deep learning methods join the research of clickbait detection. English clickbait detection corpora are built first as the foundation to perform clickbait detection. Nevertheless, public corpus based on Chinese is rarely seen making the research of clickbait detection on Chinese social media hard to continue. In this paper, a Chinese news clickbait dataset based on three mainstream Chinese media named TTNN dataset is constructed. Over 7,000 news are contained in the TTNN dataset among which over 500 news are manually labeled as clickbait based on a set of precise annotation procedures. Detailed analysis is performed on the TTNN dataset from the overall statistics, features of the title text, the source text, and the relation between title and content. The analysis shows that non-clickbait word contains more nouns and proper nouns while clickbait data contains more adjectives. Additionally, certain social medias are found to have higher clickbait-news publishing possibility. Seven classic machine learning models i.e. SVM_Linear, SVM_Sigmoid, KNN, LR, RF, DT, and ADB, and seven classic deep learning models i.e. BERT, BERT+CNN, Text-CNN, Bi-LSTM, Bi-LSTM+A, Bi-LSTM+P, Bert + Bi-LSTM+P are implemented for clickbait detection tasks as baselines. This paper proposes a deep learning model using the integration of BERT and CNN(named BCCD) achieving the best performance as high as 93.66% accuracy among other baselines ultimately. Shortcomings of the TTNN dataset and the BCCD model are also discussed in the paper. However, all analyses conducted in the paper still show the high quality and future research potential of the TTNN dataset and the BCCD model." @default.
- W4328006272 created "2023-03-22" @default.
- W4328006272 creator A5011011928 @default.
- W4328006272 creator A5020063371 @default.
- W4328006272 creator A5024620991 @default.
- W4328006272 creator A5058743657 @default.
- W4328006272 date "2022-08-01" @default.
- W4328006272 modified "2023-09-27" @default.
- W4328006272 title "Clickbait Analysis and Detection Method on Chinese Social Media" @default.
- W4328006272 cites W1832693441 @default.
- W4328006272 cites W2597366411 @default.
- W4328006272 cites W2932766729 @default.
- W4328006272 cites W2979826702 @default.
- W4328006272 cites W3120908114 @default.
- W4328006272 cites W4246826033 @default.
- W4328006272 doi "https://doi.org/10.1109/bigcom57025.2022.00049" @default.
- W4328006272 hasPublicationYear "2022" @default.
- W4328006272 type Work @default.
- W4328006272 citedByCount "0" @default.
- W4328006272 crossrefType "proceedings-article" @default.
- W4328006272 hasAuthorship W4328006272A5011011928 @default.
- W4328006272 hasAuthorship W4328006272A5020063371 @default.
- W4328006272 hasAuthorship W4328006272A5024620991 @default.
- W4328006272 hasAuthorship W4328006272A5058743657 @default.
- W4328006272 hasConcept C108583219 @default.
- W4328006272 hasConcept C12267149 @default.
- W4328006272 hasConcept C136389625 @default.
- W4328006272 hasConcept C136764020 @default.
- W4328006272 hasConcept C138885662 @default.
- W4328006272 hasConcept C154945302 @default.
- W4328006272 hasConcept C204321447 @default.
- W4328006272 hasConcept C27206212 @default.
- W4328006272 hasConcept C2776321320 @default.
- W4328006272 hasConcept C2777617010 @default.
- W4328006272 hasConcept C41008148 @default.
- W4328006272 hasConcept C50644808 @default.
- W4328006272 hasConcept C518677369 @default.
- W4328006272 hasConceptScore W4328006272C108583219 @default.
- W4328006272 hasConceptScore W4328006272C12267149 @default.
- W4328006272 hasConceptScore W4328006272C136389625 @default.
- W4328006272 hasConceptScore W4328006272C136764020 @default.
- W4328006272 hasConceptScore W4328006272C138885662 @default.
- W4328006272 hasConceptScore W4328006272C154945302 @default.
- W4328006272 hasConceptScore W4328006272C204321447 @default.
- W4328006272 hasConceptScore W4328006272C27206212 @default.
- W4328006272 hasConceptScore W4328006272C2776321320 @default.
- W4328006272 hasConceptScore W4328006272C2777617010 @default.
- W4328006272 hasConceptScore W4328006272C41008148 @default.
- W4328006272 hasConceptScore W4328006272C50644808 @default.
- W4328006272 hasConceptScore W4328006272C518677369 @default.
- W4328006272 hasLocation W43280062721 @default.
- W4328006272 hasOpenAccess W4328006272 @default.
- W4328006272 hasPrimaryLocation W43280062721 @default.
- W4328006272 hasRelatedWork W2407519794 @default.
- W4328006272 hasRelatedWork W2947809439 @default.
- W4328006272 hasRelatedWork W3093956116 @default.
- W4328006272 hasRelatedWork W3107474891 @default.
- W4328006272 hasRelatedWork W3136979370 @default.
- W4328006272 hasRelatedWork W38394648 @default.
- W4328006272 hasRelatedWork W4220838589 @default.
- W4328006272 hasRelatedWork W4244079430 @default.
- W4328006272 hasRelatedWork W4285106639 @default.
- W4328006272 hasRelatedWork W4311106074 @default.
- W4328006272 isParatext "false" @default.
- W4328006272 isRetracted "false" @default.
- W4328006272 workType "article" @default.