Marine oil spills can cause severe damage to the marine environment and biological resources. Using satellite remote sensing technology is one of the best ways to monitor the sea surface in near real-time to obtain oi...
详细信息
Marine oil spills can cause severe damage to the marine environment and biological resources. Using satellite remote sensing technology is one of the best ways to monitor the sea surface in near real-time to obtain oil spill information. The existing methods in the literature either use deep convolutional neural networks in synthetic aperture radar (SAR) images to directly identify oil spills or use traditional methods based on artificial features sequentially to distinguish oil spills from sea surface. However, both approaches currently only use image information and ignore some valuable auxiliary information, such as marine weather conditions, distances from oil spill candidates to oil spill sources, etc. In this study, we proposed a novel method to help detect marine oil spills by constructing a multi-source knowledge graph, which was the first one specifically designed for oil spill detection in the remote sensing field. Our method can rationally organize and utilize various oil spill-related information obtained from multiple data sources, such as remote sensing images, vectors, texts, and atmosphere-ocean model data, which can be stored in a graph database for user-friendly query and management. In order to identify oil spills more effectively, we also proposed 13 new dark spot features and then used a feature selection technique to create a feature subset that was favorable to oil spill detection. Furthermore, we proposed a knowledgegraph-based oil spill reasoning method that combines rule inference and graph neural network technology, which pre-inferred and eliminated most non-oil spills using statistical rules to alleviate the problem of imbalanced data categories (oil slick and non-oil slick). Entity recognition is ultimately performed on the remaining oil spill candidates using a graph neural network algorithm. To verify the effectiveness of our knowledgegraph approach, we collected 35 large SAR images to construct a new dataset, for which the training
Analyzing opinions, extracting and modeling information, and performing network analysis in online information studies are challenging tasks with multi-source social network data. This complexity arises from the diffi...
详细信息
Analyzing opinions, extracting and modeling information, and performing network analysis in online information studies are challenging tasks with multi-source social network data. This complexity arises from the difficulty in harnessing data across various platforms and the absence of a unified data modeling approach. Although social network analysis has used a multiplex approach to study complex networks, no previous work has integrated data from multiple social networks, knowledgegraph fusion, and contextual focal structure analysis (CFSA) for an online study. This study has developed a multi-sourcegraph model and applied a Cartesian merge to model relations across multiple documents, entities, and topics. We improved the information modeled with third-party data sources such as WikiData and DiffBot. This approach has created a multiplex network instance for CFSA detection, incorporating topic-topic, entity-entity, and document-document models. We applied this method to a dataset from the Indo-Pacific region and identified 40,000 unique focal sets of influential topics, entities, and documents. The top sets discussed economics, elections, and policies such as the Indo-Pacific Economic Framework, Ekonomi baru, #NKRIHargaMati, #IndonesiaJaya, and the Xinjiang Supply Chain. Our model tracks information spread across multiple social media platforms and enhances the visibility of vital information using various relationships. The results underscore the effectiveness of KG-CFSA in contextualizing large-scale information from multiple sources.
Historically, online data has provided meaningful insights for information mining, leading to the adoption of knowledgegraphs for application to online data. knowledge embedding has become an important aspect of enco...
详细信息
ISBN:
(纸本)9798400704093
Historically, online data has provided meaningful insights for information mining, leading to the adoption of knowledgegraphs for application to online data. knowledge embedding has become an important aspect of encoding and decoding links, relationships, and predicting the ties of an entity to an existing knowledgegraph. This study applied topic modeling to extract topics, entities, and themes from heterogeneous web data from different sources around the Indo-Pacific region and modeled a knowledgegraph. The knowledgegraph was subjected to knowledge embedding by applying four scoring mechanisms: ComplEx, TransE, DistMult, and HolE, on a domain knowledgegraph of Indo-Pacific Belt and Road initiatives to determine whether it was capable of revealing missing insights. This work significantly uses knowledgegraphs and embedding to understand socioeconomic-related discussions online. Valuable insights were gained from the data in this research's clustering results of knowledge embedding. Important themes such as NASAKOM and BRI were identified in Cluster 0. Cluster 1 contained themes that discussed Marxist movements synonymous with Indonesia, and Cluster 2 showed themes on China's road policies, such as Asia-Pacific Economic Cooperation and Export-Import Bank China. Cluster 3 focused mainly on China's economic policies and the Philippines. Overall, this study demonstrates the usefulness of topic modeling and knowledge embedding in uncovering insights from online data and has implications for understanding socioeconomic trends in the Indo-Pacific region.
暂无评论