武汉市初中毕业生地震知识调查

为了解初中毕业生地震科普知识的掌握情况，自行设计了武汉市初中毕业生地震科普知识调查问卷，对100名武汉市初中毕业生进行现场调查。调查结果为良、中、差等级分别占4%、18%、78%。 说明武汉市初中毕业生地震科普知识掌握情况令人堪忧，应尽早加强中学生地震科普知识教育， 以提高全民防震减灾意识。     

新型知识观下高校旅游管理专业课程知识的构建思考

旅游管理专业是1998年新设置的高校本科专业,受学科专业特点和旅游职业需求的影响,其课程知识与传统的学科专业相比,具有很强的实践性.为此,其课程知识的构建,需要深入认识旅游管理专业课程知识的特点,建立起"科学-人本观"的构建价值取向,并充分唤醒教学参与者的主体意识,发挥课堂对课程知识的构建作用.     

中生代复杂构造体系的成矿过程与成矿作用——以华北大陆北缘西拉木伦钼铜多金属成矿带为例

西拉木伦钼铜多金属成矿带处于华北克拉通与中亚造山带的过渡区,是古生代古亚洲构造域与中生代西太平洋构造域的交汇部位。在中生代受多种构造体系的制约,如中亚造山带造山后期局部伸展、蒙古-鄂霍茨克俯冲-碰撞造山作用、古太平洋板块的向西俯冲和中国东部岩石圈减薄事件的影响等。西拉木伦成矿带成矿斑岩锆石U-Pb年龄和辉钼矿Re-Os同位素年龄资料显示,钼铜矿成岩成矿主要集中在260～220Ma、180～150Ma和140～120Ma三个时期。结合华北克拉通北缘构造演化历史,推测这三期成矿作用主要与造山后局部伸展、构造体系转折和陆内伸展(岩石圈减薄)过程有关,并相应建立了"车户沟式"、"鸡冠山式"和"敖伦花式"三类斑岩钼铜矿床成矿模式。进一步研究表明,岩石的酸碱性、岩浆来源、岩浆的氧逸度、岩浆演化方式、构造背景等因素,制约了成矿作用的专属性。     

冰雪灾害条件下地质灾害应急抢险处置知识构成分析

知识库是冰雪灾害条件下地质灾害应急抢险决策支持系统的核心,知识库中的知识构成则是整个研究的基础。本文按照产生式规则法的知识表示方式,分别分析了冰雪灾害条件下地质灾害应急抢险处置知识的前提及结论部分的主要构成内容,并总结了应急抢险处置措施选择应该遵循的基本原则,另外还给出了根据知识构成抽取具体知识的范例,为冰雪灾害条件下地质灾害应急抢险知识库及决策支持系统的建立提供了基础性研究成果。     

基于ALBERT模型的园林植物知识实体与关系抽取方法

园林植物知识图谱可为顾及区域适应性、观赏性和生态性等因子的绿化树种的选型提供知识支持。植物描述文本的实体识别及关系抽取是知识图谱构建的关键环节。针对植物领域未有公开的标注数据集,本文阐述了园林植物数据集的构建流程,定义了园林植物的概念体系结构,完成了园林植物语料库的构建。针对现有Word2vec、ELMo和BERT等语言模型存在无法解决多义词、融合上下文能力差、运行速度慢等缺点,提出了嵌入ALBERT（A Lite BERT）预训练语言模型的实体识别和关系抽取模型。ALBERT预训练的动态词向量能够有效地表示文本特征,将其分别输入到BiGRU-CRF命名实体识别模型和BiGRU-Attention关系抽取模型中进行训练,进一步提升实体识别和关系抽取的效果。在园林植物语料库上进行方法的有效性验证,结果表明ALBERT-BiGRU-CRF命名实体识别模型的F₁值为0.9517,ALBERT-BiGRU-Attention关系抽取模型的F₁值为0.9161,相较于经典的语言模型（如Word2vec、ELMo和BERT等）性能有较为显著的提升。因此基于ALBERT模型的实体与关系抽取任务能有效提高识别分类效果,可将其应用于植物描述文本的实体关系抽取任务中,为园林植物知识图谱自动构建提供方法。     

协同多时相波谱特征的不透水面信息级联提取

不透水面作为反应城市表征变化和区域城镇化的重要技术指标,其位置、图斑大小、空间分布等信息在地表水热循环和能量平衡等领域被广泛需求。传统方法大都基于单一时相信息提取不透水面,而忽略多时相所蕴含的丰富信息。因此,本文提出多时相信息融合的不透水面级联提取方法,利用Landsat-8 OLI遥感影像分析归一化植被指数（Normalized Difference Vegetation Index, NDVI）、改进的归一化水体指数（Modified Normalized Difference Water Index, MNDWI）和归一化建筑指数（Normalized Difference Building Index, NDBI）年内时序变化特点和典型地物间多时相波谱曲线的协同特征,并归纳不透水面多时相变化规律;再根据先验知识所获取的有效地表信息,进行多时相分级提取不透水面信息。此外,基于实地考察数据和同期2 m GF-1遥感影像屏幕数字化生成30 m不透水面图斑,进行精度验证、分析和对比单时相、四季相及多时相3种时序情况下的提取精度。结果表明：单时相提取不透水面总精度最低,四季相提取精度优于单时相,而多时相提取精度最高（精度可达93.66%,Kappa系数为0.81）。本方法在偏远城镇不透水面的有效识别中显露潜在优势,可为不透水面提取方法融合时序波谱特征提供新思路。     

The Value of Practice-Based Knowledge

Increasing recognition of the value of practice-based or experiential knowledge in natural resource management justifies the creation of a new category of articles in Society & Natural Resources that we are calling Practice-Based Knowledge (PBK). The rationale for focusing on PBK is due to its key role in the emergence of hybrid governance institutions across state, market, and civil society, understanding the complexity of dynamic socioecological systems, recognizing the challenges of multiple knowledge systems and context-specific practices, embracing the power of informal institutions and civic science, and engaging debates on the growing prevalence of market-oriented conservation. The goal is to provide a dedicated space within the published, peer-reviewed literature for scholars, government officials, nonprofit managers, and engaged citizens to share experiences informed by practical action. Relevant and timely practice-based insights may improve understanding and management of social and ecological processes and systems, while also offering the potential to contribute to theory.     

In Situ Maintenance of Traditional Ecological Knowledge on Malekula Island,Vanuatu

Traditional ecological knowledge (TEK) guides resource management across the globe, but is at risk amid social and ecological change. This has prompted numerous calls for TEK maintenance efforts, but these remain largely unexamined in the literature. Here, we discuss three examples of in situ TEK maintenance from Malekula Island in Vanuatu, locally known as kastom schools. Based on qualitative data, we find that the kastom schools may create several opportunities to maintain TEK (e.g., establishing local control over education), and argue that they represent the creative and adaptive management of tradition in dynamic social–ecological contexts. However, a number of challenges, both practical (e.g., lack of funding) and epistemological (e.g., changing modes of cultural transmission), threaten the efficacy of the kastom schools. We argue that in situ modes of TEK maintenance have promise, but that issues of power and heterogeneity require serious consideration if such measures are to succeed.     

Modeling and representation for earthquake emergency response knowledge: perspective for working with geo-ontology

Earthquake emergency response is one of the three earthquake disaster mitigation work systems in China, already achieving good results in some earthquake disaster mitigation situations. Earthquake emergency plans and emergency command systems are among the most important research and operations components of emergency response. These components commonly come with challenges, such as the pertinence of emergency commands and the operability of the countermeasures to be improved. The promise for solving this problem resides with applying knowledge that aids intelligence creation for decision-making. In this paper, we put forward a conceptual model of knowledge for earthquake disaster emergency response (EDER); compositions of EDER knowledge are introduced within architecture. A modeling method incorporating geo-ontology is used to build basic modeling primitives. Geo-ontology serves to represent geospatial characteristics of the EDER knowledge and addresses a need for semantic interoperability in the modeling process. A decision problem framework and a case study have been used as theoretical framework and an application test, respectively, to evaluate the EDER knowledge architecture and models. The EDER knowledge model provides a foundation for intelligent emergency response that helps solve knowledge problems to improve earthquake disaster response.     

Making a Mark: a computational and visual analysis of one researcher’s intellectual domain

Dr. David Mark is widely regarded as a path-breaking researcher in geographic information science. What are the structural and temporal characteristics of his intellectual contributions, as seen through the eyes of the broader academic community? Aiming to answer that question, this article presents a scientometric analysis of publications that have been cited alongside David Mark’s papers. In deliberate contrast to the widespread focus on using citation data to condense scientific impact into a handful of indicators, the methodological contribution of this study lies in its mix of computational and visualization approaches. In the search for latent domain structures, state-of-the-art practices in information science, bibliometrics, and network visualization are combined and extended. An initial network of 50,000+ publications and 4,000,000+ document co-citations undergoes a series of transformations reducing it to 9000 publications that are then clustered in a two-stage process, leading to 678 communities whose co-citation linkages are used to delineate 19 super-communities. To enable replication of this approach for other studies, much focus in this article is on detailed discussion of that workflow as well as on highlighting the reasoning behind the choices made among data sources and analytical methods. The topical evolution of David Mark’s domain of influence is explored in some detail, based on tabular and graphic representations of extracted community structures. Results confirm not only the enormous overall breadth of his influence but also how lasting and recurrent it has been in some areas.