首页|Study Results from University of Ljubljana Provide New Insights into Artificial Intelligence (Towards Improved Knowledge About Waterrelated Extremes Based On News Media Information Captured Using Artificial Intelligence)
Study Results from University of Ljubljana Provide New Insights into Artificial Intelligence (Towards Improved Knowledge About Waterrelated Extremes Based On News Media Information Captured Using Artificial Intelligence)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Elsevier
Fresh data on Artificial Intelligence are presented in a new report. According to news originating from Ljubljana, Slovenia, by NewsRx correspondents, research stated, “Recent advances in machine learning have enabled near real-time retrieval of information from textual documents impacting a wide range of knowledge domains. This advantage makes of the insight extracted from text-based documents (e.g., reports and news) on water-related events an invaluable source of information that complements traditional approaches.” Funders for this research include European Commission Joint Research Centre, Slovenian Research and Innovation Agency, University of Ljubljana by Development Fund for UNESCO Chair on Water-related Disaster Risk Reduction, European Research Council (ERC). Our news journalists obtained a quote from the research from the University of Ljubljana, “By leveraging machine learning, we can not only characterize the event and determine its magnitude or phases of extreme weather event, but also identify its core elements. This is especially crucial in the current era of climate change, where extreme weather events such as floods or thunderstorms are becoming more frequent and unpredictable. By improving our ability to detect and analyse such events, we can enhance our alert systems and take more effective action to mitigate their impact. In this paper we discuss the role of worldwide media observation in extracting and estimating hydrological characteristics of floods, droughts, and heat waves, through the analysis of three case studies, complementing the information provided by traditional monitoring and measurement methods as an earlier but weaker signal. The results presented in this study indicate that the news media signal can be regarded as relatively good proxy of flood dynamics. It can capture the temporal dynamics of the event, and, in some cases, there could be a clear up to 1-day lag between the peak discharge values (i.e., the most extreme flood situation) and the peak in the number of published news. This lag can be attributed to the time needed by journalists to respond to the situation in publishing related news articles covering the event. Our result show that national and regional news can cover well local events. When compared to floods, drought conditions are less explicitly detected from the media. Our result show that European April 2022 drought did not produce much activity in the media while the combination of drought and extreme heat in July 2022 yielded a significant media coverage throughout the Europe. Hence, this can be attributed to the fact that hydrological drought such as low river flows do not attract much attention by the media unless there is a significant impact on the society.”
LjubljanaSloveniaEuropeArtificial IntelligenceCyborgsEmerging TechnologiesMachine LearningUniversity of Ljubljana
NETL
NSTL
Elsevier
