An improved deep belief neural network based civil unrest event forecasting in twitter

Iyda, J. Joslin; Geetha, P.

doi:10.1007/s10489-022-03746-3

Cited by 2 publications

(1 citation statement)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Twitter serves as a widely used social media platform that contains abundant information regarding planned events. Iyda and Geetha [15] introduced an Improved Deep Belief Neural Network (iDBNN) to predict protests using Twitter data, where the efficiency of the proposed method was validated with the case study of the 2019 Hong Kong protests. In addition to Twitter, Google Trends (GT) offers a valuable gateway to access extensive big data on various global topics.…”

Section: Planned Event Predictionmentioning

confidence: 99%

Exploring and Visualizing Research Progress and Emerging Trends of Event Prediction: A Survey

Xu,

Liu,

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Over the last decade, event prediction has drawn attention from both academic and industry communities, resulting in a substantial volume of scientific papers published in a wide range of journals by scholars from different countries and disciplines. However, thus far, a comprehensive and systematic survey of recent literature has been lacking to quantitatively capture the research progress as well as emerging trends in the event prediction field. Aiming at addressing this gap, we employed CiteSpace software to analyze and visualize data retrieved from the Web of Science (WoS) database, including authors, documents, research institutions, and keywords, based on which the author co-citation network, document co-citation network, collaborative institution network, and keyword co-occurrence network were constructed. Through analyzing the aforementioned networks, we identified areas of active research, influential literature, collaborations at the national level, interdisciplinary patterns, and emerging trends by identifying the central nodes and the nodes with strong citation bursts. It reveals that sensor data has been widely used for predicting weather events and meteorological events (e.g., monitoring sea surface temperature and weather sensor data for predicting El Nino). The real-time and multivariable monitoring features of sensor data enable it to be a reliable source for predicting multiple types of events. Our work offers not only a comprehensive survey of the existing studies but also insights into the development trends within the event prediction field. These findings will assist researchers in conducting further research in this area and draw a large readership among academia and industrial communities who are engaged in event prediction research.

show abstract

Section: Planned Event Predictionmentioning

confidence: 99%