There are mainly four kinds of models to record and deal with historical information.By taking them as reference,the spatio-temporal model based on event semantics is proposed.In this model,according to the way for de...There are mainly four kinds of models to record and deal with historical information.By taking them as reference,the spatio-temporal model based on event semantics is proposed.In this model,according to the way for describing an event,all the information are divided into five domains.This paper describes the model by using the land parcel change in the cadastral information system,and expounds the model by using five tables corresponding to the five domains.With the aid of this model,seven examples are given on historical query,trace back and recurrence.This model can be implemented either in the extended relational database or in the object-oriented database.展开更多
Predicting information dissemination on social media,specifcally users’reposting behavior,is crucial for applications such as advertising campaigns.Conventional methods use deep neural networks to make predictions ba...Predicting information dissemination on social media,specifcally users’reposting behavior,is crucial for applications such as advertising campaigns.Conventional methods use deep neural networks to make predictions based on features related to user topic interests and social preferences.However,these models frequently fail to account for the difculties arising from limited training data and model size,which restrict their capacity to learn and capture the intricate patterns within microblogging data.To overcome this limitation,we introduce a novel model Adapt pre-trained Large Language model for Reposting Prediction(ALL-RP),which incorporates two key steps:(1)extracting features from post content and social interactions using a large language model with extensive parameters and trained on a vast corpus,and(2)performing semantic and temporal adaptation to transfer the large language model’s knowledge of natural language,vision,and graph structures to reposting prediction tasks.Specifcally,the temporal adapter in the ALL-RP model captures multi-dimensional temporal information from evolving patterns of user topic interests and social preferences,thereby providing a more realistic refection of user attributes.Additionally,to enhance the robustness of feature modeling,we introduce a variant of the temporal adapter that implements multiple temporal adaptations in parallel while maintaining structural simplicity.Experimental results on real-world datasets demonstrate that the ALL-RP model surpasses state-of-the-art models in predicting both individual user reposting behavior and group sharing behavior,with performance gains of 2.81%and 4.29%,respectively.展开更多
Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features o...Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features of its grammatical categories have been neglected.This paper investigates into the categorization of zài by conducting a morphosyntactic test among College English majors in China.The results show that:prototypes are organizing the grammatical categories of zài at all levels in terms of intra-categorial gradience;the semantic construal of zài construction could significantly influence the accuracy of the grammatical categorization of zài;the syntactic structure can provide viable cue for the identification of grammatical categories of zài;spatiality,temporality and the status of existing are three essential semantic features encoded by zài,the concurrence of which leads to various degree of inter-categorial vagueness,indicating a conflict between the rigid grammatical classification and the indeterminate nature of the grammatical functions of zai,suggesting the necessity to reconsider the efficacy of applying indiscriminately the Anglo-Saxon grammar into the study of Chinese spatial-temporal constructions.展开更多
现有的下一个兴趣点(point of interest,PoI)推荐技术存在三个主要问题:使用过于简单的方法构建用户兴趣模型、忽略用户和PoI之间在时空维度上的互动以及未能充分挖掘用户间复杂的高阶交互信息。针对这些问题,提出一种新颖的超图学习模...现有的下一个兴趣点(point of interest,PoI)推荐技术存在三个主要问题:使用过于简单的方法构建用户兴趣模型、忽略用户和PoI之间在时空维度上的互动以及未能充分挖掘用户间复杂的高阶交互信息。针对这些问题,提出一种新颖的超图学习模型FSTMH,细粒度地融合时间、空间和语义信息,用于下一个PoI推荐。FSTMH包括细粒度嵌入模块和多层次嵌入模块。前者通过使用地理图卷积网络和有向超图卷积网络进行学习,获取对应的嵌入信息,并通过对比学习提升PoI表示的质量,使用细粒度超图卷积网络学习该模块的PoI嵌入;后者将多层语义超图输入到多层超图卷积网络,学习多层次语义的PoI嵌入表示。最后,模型将两个模块的PoI嵌入向量进行组合,生成最终的top-K预测结果。通过在广泛使用的三个社交网络公共数据集上进行多种实验,结果均表明FSTMH模型表现出色,说明该新模型可作为提高下一个PoI推荐的有效方法。展开更多
[目的]现有的语义变化检测方法对于遥感影像的局部和全局特征利用不足,且忽略了不同时相间的时空依赖性,导致土地覆盖语义分类结果不精确。此外,检测的变化对象存在边缘模糊问题,检测结果和实际变化区域的一致性有待提升。[方法]针对这...[目的]现有的语义变化检测方法对于遥感影像的局部和全局特征利用不足,且忽略了不同时相间的时空依赖性,导致土地覆盖语义分类结果不精确。此外,检测的变化对象存在边缘模糊问题,检测结果和实际变化区域的一致性有待提升。[方法]针对这些挑战,受具有长序列处理能力的视觉状态空间模型(Vision State Space Model, VSSM)启发,本文提出了一种融合卷积神经网络(Convolutional Neural Networks, CNN)与VSSM的语义变化检测网络CVS-Net。该网络有效结合了CNN的局部特征提取优势与VSSM捕捉长距离依赖关系的能力,并嵌入基于VSSM的双向时空关系建模模块以引导网络深入理解影像间的时空变化逻辑关系。此外,为增强模型对变化对象边缘的识别精度,提出了边缘感知强化分支,通过联合拉普拉斯算法和多任务架构自动集成边界信息,增强模型对于变化地物的形状模式的学习能力。[结果]在SECOND和FZSCD数据集上,将本方法与HRSCD.str4、Bi-SRNet、ChangeMamba、ScanNet及TED五种主流的SCD方法进行对比。实验结果显示,本方法的语义变化检测性能优于其他对比方法,验证了本方法的有效性。在SECOND数据集上实现了23.95%的分离卡帕系数(Sek)和72.89%的平均交并比(mIoU),在FZ-SCD数据集上的SeK达到23.02%, mIoU达到72.60%。消融实验结果中,随着在基础网络中添加各模块,SeK值逐步提升至21.26%、23.04%和23.95%,证明了CVS-Net中各模块的有效性。[结论]本方法可有效提升了语义变化检测的属性和几何精度,可为城市可持续发展、土地资源管理等应用领域提供技术参考和数据支撑。展开更多
土地覆盖分类体系是土地覆盖研究中的重要内容。该文总结了9种主要土地覆盖分类体系,展示了分类体系和相应数据产品,分析了体系之间的区别和联系,讨论了土地覆盖分类体系精细程度,以及分类体系、空间分辨率、空间覆盖范围的关系,讨论了...土地覆盖分类体系是土地覆盖研究中的重要内容。该文总结了9种主要土地覆盖分类体系,展示了分类体系和相应数据产品,分析了体系之间的区别和联系,讨论了土地覆盖分类体系精细程度,以及分类体系、空间分辨率、空间覆盖范围的关系,讨论了分类体系之间的语义一致性。论文认为土地覆盖分类体系(land cover classification system, LCCS)和全球土地覆盖精细分辨率观测与监测数据(finer resolution observation and monitoring of global land cover, FROM-GLC)分类体系在精细化分类方面具有优势,高空间分辨率精细分类存在较大技术挑战和实现难度;目前分类体系之间在逻辑关系、精细分类、名称定义、代码等多方面,存在明显的语义不一致现象;最后总结指出全球土地覆盖分类研究存在全球化和区域化并存、分类更加精细、产品精度更高、时间间隔和空间分辨率更加细致的发展趋势,数据产品的语义不一致性还需改进,未来需在加强分类体系的兼容性、实现数据产品的共享互操作方面提出解决方案。展开更多
文摘There are mainly four kinds of models to record and deal with historical information.By taking them as reference,the spatio-temporal model based on event semantics is proposed.In this model,according to the way for describing an event,all the information are divided into five domains.This paper describes the model by using the land parcel change in the cadastral information system,and expounds the model by using five tables corresponding to the five domains.With the aid of this model,seven examples are given on historical query,trace back and recurrence.This model can be implemented either in the extended relational database or in the object-oriented database.
文摘Predicting information dissemination on social media,specifcally users’reposting behavior,is crucial for applications such as advertising campaigns.Conventional methods use deep neural networks to make predictions based on features related to user topic interests and social preferences.However,these models frequently fail to account for the difculties arising from limited training data and model size,which restrict their capacity to learn and capture the intricate patterns within microblogging data.To overcome this limitation,we introduce a novel model Adapt pre-trained Large Language model for Reposting Prediction(ALL-RP),which incorporates two key steps:(1)extracting features from post content and social interactions using a large language model with extensive parameters and trained on a vast corpus,and(2)performing semantic and temporal adaptation to transfer the large language model’s knowledge of natural language,vision,and graph structures to reposting prediction tasks.Specifcally,the temporal adapter in the ALL-RP model captures multi-dimensional temporal information from evolving patterns of user topic interests and social preferences,thereby providing a more realistic refection of user attributes.Additionally,to enhance the robustness of feature modeling,we introduce a variant of the temporal adapter that implements multiple temporal adaptations in parallel while maintaining structural simplicity.Experimental results on real-world datasets demonstrate that the ALL-RP model surpasses state-of-the-art models in predicting both individual user reposting behavior and group sharing behavior,with performance gains of 2.81%and 4.29%,respectively.
文摘Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features of its grammatical categories have been neglected.This paper investigates into the categorization of zài by conducting a morphosyntactic test among College English majors in China.The results show that:prototypes are organizing the grammatical categories of zài at all levels in terms of intra-categorial gradience;the semantic construal of zài construction could significantly influence the accuracy of the grammatical categorization of zài;the syntactic structure can provide viable cue for the identification of grammatical categories of zài;spatiality,temporality and the status of existing are three essential semantic features encoded by zài,the concurrence of which leads to various degree of inter-categorial vagueness,indicating a conflict between the rigid grammatical classification and the indeterminate nature of the grammatical functions of zai,suggesting the necessity to reconsider the efficacy of applying indiscriminately the Anglo-Saxon grammar into the study of Chinese spatial-temporal constructions.
文摘现有的下一个兴趣点(point of interest,PoI)推荐技术存在三个主要问题:使用过于简单的方法构建用户兴趣模型、忽略用户和PoI之间在时空维度上的互动以及未能充分挖掘用户间复杂的高阶交互信息。针对这些问题,提出一种新颖的超图学习模型FSTMH,细粒度地融合时间、空间和语义信息,用于下一个PoI推荐。FSTMH包括细粒度嵌入模块和多层次嵌入模块。前者通过使用地理图卷积网络和有向超图卷积网络进行学习,获取对应的嵌入信息,并通过对比学习提升PoI表示的质量,使用细粒度超图卷积网络学习该模块的PoI嵌入;后者将多层语义超图输入到多层超图卷积网络,学习多层次语义的PoI嵌入表示。最后,模型将两个模块的PoI嵌入向量进行组合,生成最终的top-K预测结果。通过在广泛使用的三个社交网络公共数据集上进行多种实验,结果均表明FSTMH模型表现出色,说明该新模型可作为提高下一个PoI推荐的有效方法。
文摘[目的]现有的语义变化检测方法对于遥感影像的局部和全局特征利用不足,且忽略了不同时相间的时空依赖性,导致土地覆盖语义分类结果不精确。此外,检测的变化对象存在边缘模糊问题,检测结果和实际变化区域的一致性有待提升。[方法]针对这些挑战,受具有长序列处理能力的视觉状态空间模型(Vision State Space Model, VSSM)启发,本文提出了一种融合卷积神经网络(Convolutional Neural Networks, CNN)与VSSM的语义变化检测网络CVS-Net。该网络有效结合了CNN的局部特征提取优势与VSSM捕捉长距离依赖关系的能力,并嵌入基于VSSM的双向时空关系建模模块以引导网络深入理解影像间的时空变化逻辑关系。此外,为增强模型对变化对象边缘的识别精度,提出了边缘感知强化分支,通过联合拉普拉斯算法和多任务架构自动集成边界信息,增强模型对于变化地物的形状模式的学习能力。[结果]在SECOND和FZSCD数据集上,将本方法与HRSCD.str4、Bi-SRNet、ChangeMamba、ScanNet及TED五种主流的SCD方法进行对比。实验结果显示,本方法的语义变化检测性能优于其他对比方法,验证了本方法的有效性。在SECOND数据集上实现了23.95%的分离卡帕系数(Sek)和72.89%的平均交并比(mIoU),在FZ-SCD数据集上的SeK达到23.02%, mIoU达到72.60%。消融实验结果中,随着在基础网络中添加各模块,SeK值逐步提升至21.26%、23.04%和23.95%,证明了CVS-Net中各模块的有效性。[结论]本方法可有效提升了语义变化检测的属性和几何精度,可为城市可持续发展、土地资源管理等应用领域提供技术参考和数据支撑。
文摘土地覆盖分类体系是土地覆盖研究中的重要内容。该文总结了9种主要土地覆盖分类体系,展示了分类体系和相应数据产品,分析了体系之间的区别和联系,讨论了土地覆盖分类体系精细程度,以及分类体系、空间分辨率、空间覆盖范围的关系,讨论了分类体系之间的语义一致性。论文认为土地覆盖分类体系(land cover classification system, LCCS)和全球土地覆盖精细分辨率观测与监测数据(finer resolution observation and monitoring of global land cover, FROM-GLC)分类体系在精细化分类方面具有优势,高空间分辨率精细分类存在较大技术挑战和实现难度;目前分类体系之间在逻辑关系、精细分类、名称定义、代码等多方面,存在明显的语义不一致现象;最后总结指出全球土地覆盖分类研究存在全球化和区域化并存、分类更加精细、产品精度更高、时间间隔和空间分辨率更加细致的发展趋势,数据产品的语义不一致性还需改进,未来需在加强分类体系的兼容性、实现数据产品的共享互操作方面提出解决方案。
