基于混合过滤的地学数据个性化推荐方法设计与实现

推荐系统是帮助互联网用户克服信息过剩的有效工具。在地学数据共享领域,较其他物品的内容属性,地学数据具有更加丰富的时空属性,这也给地学数据推荐带来挑战。针对地学数据的特点,为地学数据共享推荐服务开发了一种动态加权的混合过滤方法。该方法分别采用协同过滤和基于内容过滤算法预测用户对数据的兴趣度,再以训练模型计算最优加权权重,计算最终预测评分。在数据获取阶段,通过用户访问日志数据,采用Jenks Natural Break算法分析用户访问记录获取用户的数据兴趣度。在基于内容过滤部分,通过数据的空间、时间及内容属性计算数据相似度,并以用户历史行为为依据计算用户兴趣。在协同过滤和基于内容过滤中分别采用k-NN算法计算用户对未访问数据的预测评分,并进行加权求和。通过训练集,对理想权重值及用户的共同评价度（co-rating level）进行建模,拟合二者的关系。该模型被应用于混合过滤的权重调整,以获得最优的加权方程。测试结果显示,结合数据时空属性的混合过滤方法的准确度和召回率,较单一的协同过滤或基于内容过滤方法有显著提高。     

基于知识的遥感图像分类方法的探讨

本文提出了一种基于知识的遥感图像土地利用分类法。包括知识的获取和表达 ,以及将知识用于土地利用分类。知识的获取和表达包括两方面。一是将原有的GIS土地利用数据与新的遥感图像配准、叠加、集成 ,并从中挖掘各类型的光谱知识。二是通过对研究区各土地利用类型变化特点进行分析研究 ,并从中得到原有各土地利用类型经数年后变成其它类型的可能性及其大小。并将其定量化表达成新遥感图像上各类型的先验性概率。在知识的应用和分类方面 ,分别对原土地利用各类型所对应的遥感图像进行知识的应用和处理 ,利用各自可能出现的类型及其各类型的先验性概率进行最大似然法分类。研究表明 ,该方法将原有的GIS数据和各种土地利用类型变化的先验性知识综合集成用于新的遥感图像的分类中。这不仅可以促进GIS数据更新的自动化 ,而且还可以得到比常规最大似然法高的分类精度。     

GIS空间关联规则挖掘研究

提取不同对象或现象间相互联系的空间规律,发现、解释或预测空间现象或事件,是GIS空间关联规则挖掘价值的体现。通过对目前的空间关联规则挖掘理论和实例的研究,分析了挖掘过程涉及到的基本问题：挖掘实施的基本步骤、开发路线、挖掘有效性的保证,挖掘方法及优化策略等,得到了实用的挖掘的一般方法,也对挖掘的应用实例进行了分析,展望了应用前景。     

Expanded adaptive fuzzy sliding mode controller using expert knowledge and fuzzy basis function expansion for UFV depth control

Generally, the underwater flight vehicle (UFV) depth control system operates with the following problems: it is a multi-input multi-output (MIMO) system, it requires robustness, a continuous control input, and further, it has the speed dependency of controller parameters. To solve these problems, an expanded adaptive fuzzy sliding mode controller (EAFSMC), which is based on the decomposition method designed by using an expert knowledge and the decoupled sub-controllers and composition method designed by using the fuzzy basis function expansions (FBFEs), is proposed. To verify the performance of the EAFSMC, the depth control of UFV in various operating conditions is performed. Simulation results show that the EAFSMC solves all problems experienced in the UFV depth control system online.     

培育志趣，引导学生奔向地球科学大道

随着人类社会和科学技术的发展,尤其是中国国力的提升和教育的深化提高,教育改革,包括地学高等教育的改革,就更需进一步深化,以适应时代发展的需要。文章根据作者长期从事地学教学实践与科研的经历,针对2020年全国基础地质学教学会议关于普通地质学教学改革的精神,概括提出两点体会,一是关于普地教学目标和任务,二是涉及普地教学关键所在,供同行讨论。作者从地球科学基本特征和教学目标出发,提出普地教学的核心价值与重要性,进一步明确普地教学的关键在于教师队伍、教学定位与目标及教学内容与方法等三个方面。普通地质学的教学是地学高等教育中最基础、最关键的环节,其总体目标是提高地学教育的普适性,目的是培育学生的科学志趣,引导学生奔向地球科学大道。     

A Decision Support System for the Evaluation of Eco-engineering Strategies for Slope Protection

A decision support system (DSS) has been developed to assist expert and non-expert users in the evaluation and selection of eco-engineering strategies for slope protection. This DSS combines a qualitative hazard assessment of erosion and mass movements with a detailed catalogue of eco-engineering strategies for slope protection of which the suitability is evaluated in relation to the data entered. The slope decision support system (SDSS) is a knowledge based DSS in which knowledge is stored in frames containing rules that can evaluate the available information for a project, stored as project specific information (PSI) in a data file. The advantages of such a system are that it accepts incomplete information and that the qualitative nature of the information does not instil the user with a sense of unjustified exactitude. By its multidisciplinary and progressive nature, the DSS will be of value during the initial stages of an eco-engineering project when data collection and the potential of different eco-engineering strategies are considered. The accent of the output of the DSS is on the application of eco-engineering strategies for slope protection as an environmentally-friendly solution aiding sustainable development. For its acceptance within the engineering community, the DSS needs to prove its predictive capacity. Therefore, its performance has been benchmarked against successful and unsuccessful cases of slope stabilisation using eco-engineering. The target audience and the areas of application of this DSS are reviewed and the strategies for further development in this area suggested.     

上海天文台2004年度的科研工作

简要回顾了中国科学院上海天文台2004年科研及相关工作情况。     

Science Requirements on Future Missions and Simulated Mission Scenarios

The science requirements on future gravity satellite missions, following from the previous contributions of this issue, are summarized and visualized in terms of spatial scales, temporal behaviour and accuracy. This summary serves the identification of four classes of future satellite mission of potential interest: high-altitude monitoring, satellite-to-satellite tracking, gradiometry, and formation flights. Within each class several variants are defined. The gravity recovery performance of each of these ideal missions is simulated. Despite some simplifying assumptions, these error simulations result in guidelines as to which type of mission fulfils which requirements best.     

动态网页数据库教学方法探索

以任务教学法和难题教学法为理论基础,研究了在学科教学当中,如何用最有效的方法帮助学生建立动态网页数据库的知识体系.从实践层面上分析了网页数据库教学的基本特点,并提出了该课程教学的若干原则.在实践层面上,为同行工作者提供了一个可供借鉴的课程模型,最终达到培养学生的兴趣和能力的双重教学目的.     

REVIEWS OF GEOGRAPHIC SOFTWARE

Geographic Software Reviewed in this article: LIFE TABLES and the LESLIE MATRIX. CONDUIT MENTMAP2 . Lawrence W. Carstensen Jr . PC-MATLAB , Version 1.0. The Math Works, Inc. QUICKMAP , version 1.0. Sammamish Data Systems, Inc. SYSTAT: THE SYSTEM FOR STATISTICS , Version 2.1, Systat Inc. USA DISPLAY . Instant Tecall