Building a model-based personalised recommendation approach for tourist attractions from geotagged social media data     

Xiaoyu Sun  Xia Peng  Yiran Chen  Yu Liu 《International Journal of Digital Earth》2019,12(6):661-678

When travelling, people are accustomed to taking and uploading photos on social media websites, which has led to the accumulation of huge numbers of geotagged photos. Combined with multisource information (e.g. weather, transportation, or textual information), these geotagged photos could help us in constructing user preference profiles at a high level of detail. Therefore, using these geotagged photos, we built a personalised recommendation system to provide attraction recommendations that match a user's preferences. Specifically, we retrieved a geotagged photo collection from the public API for Flickr (Flickr.com) and fetched a large amount of other contextual information to rebuild a user's travel history. We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation (the matching process) and candidate ranking. In the matching process, we used a support vector machine model that was modified for multiclass classification to generate the candidate list. In addition, we used a gradient boosting regression tree to score each candidate and rerank the list. Finally, we evaluated our recommendation results with respect to accuracy and ranking ability. Compared with widely used memory-based methods, our proposed method performs significantly better in the cold-start situation and when mining ‘long-tail’ data.  相似文献   

基于超像元词包特征和主动学习的高分遥感影像变化检测     

杨进一  徐伟铭  王成军  翁谦 《地球信息科学学报》2019,21(10):1594-1607

为解决高分辨率遥感影像变化检测中存在底层特征缺乏语义信息、像元级的检测结果存在“椒盐”现象以及监督分类中样本标注自动化程度较低,本文提出一种基于超像元词包特征和主动学习的变化检测方法。首先采用熵率分割算法获取叠加影像的超像元对象;其次提取两期影像像元点对间的邻近相关影像特征（相关度、斜率和截距）和顾及邻域的纹理变化强度特征（均值、方差、同质性和相异性）,经线性组合作为像元点对的底层特征;然后基于像元点对底层特征利用BOW模型构建超像元词包特征,并采用一种改进标注策略的主动学习方法从无标记样本池中优选信息量较大的样本,且自动标注样本类别;最后训练分类器模型完成变化检测。通过选用2组不同地区的GF-2影像和Worldview-Ⅱ影像作为数据源进行实验,实验结果中2组数据集的F1分数分别为0.8714、0.8554,正确率分别为0.9148、0.9022,漏检率分别为0.1681、0.1868,误检率分别为0.0852、0.0978。结果表明,该法能有效识别变化区域、提高变化检测精度。此外,传统主动学习方法与改进标注策略的主动学习方法的学习曲线对比显示,改进的标注策略可在较低精度损失下,有效提高样本标注自动化程度。  相似文献   

Dissolved organic matter in surface runoff in the Loess Plateau of China: The role of rainfall events and land-use     

Zhongwu Li  Hao Peng  Binggeng Xie  Chun Liu  Xiaodong Nie  Danyang Wang  Mei Huang  Haibing Xiao  Lin Shi  Xuqin Zhang  Jieyu Jiang 《水文研究》2020,34(6):1446-1459

Exploring the chemical characterization of dissolved organic matter (DOM) is important for understanding the fate of laterally transported organic matter in watersheds. We hypothesized that differences in water-extractable organic matter (WEOM) in soils of varying land uses and rainfall events may significantly affect the quality and the quantity of stream DOM. To test our hypotheses, characteristics of rainfall-runoff DOM and WEOM of source materials (topsoil from different land uses and gullies, as well as typical vegetation) were investigated at two adjacent catchments in the Loess Plateau of China, using ultraviolet–visible absorbance and excitation emission matrix fluorescence with parallel factor analysis (PARAFAC). Results indicated that land-use types may significantly affect the chemical composition of soil WEOM, including its aromaticity, molecular weight, and degree of humification. The PARAFAC analysis demonstrated that the soils and stream water were dominated by terrestrial/allochthonous humic-like substances and microbial transformable humic-like fluorophores. Shifts in the fluorescence properties of stream DOM suggested a pronounced change in the relative proportion of allochthonous versus autochthonous material under different rainfall patterns and land uses. For example, high proportions of forestland could provide more allochthonous DOM input. This study highlights the relevance of soils and hydrological dynamics on the composition and fluxes of DOM issuing from watersheds. The composition of DOM in soils was influenced by land-use type. Precipitation patterns influenced the proportion of terrestrial versus microbial origins of DOM in surface runoff. Contributions of allochthonous, terrestrially derived DOM inputs were highest from forested landscapes.  相似文献   

多层砂土地基扩底桩单桩抗压模型试验及颗粒流模拟研究     

李飞  杨俊杰  宋琦  孙涛 《中国海洋大学学报(自然科学版)》2020,(2):116-125

利用室内半模试验和颗粒流数值模拟,揭示多层砂土地基扩底桩单桩抗压承载特性及变形特征。结果表明,通过对比分析极限承载力与H_h/D(持力层厚度与扩大头直径之比)的关系可以看出,单桩的抗压极限承载力随H_h/D逐渐增加,当H_h/D超过2.0时,极限承载力基本不再增加,此时的单桩抗压极限承载力稳定在300.01~303.25 N,是H_h/D=0.5时极限承载力(183.83 N)的1.65倍。扩大头下部土体发生局部压缩-剪切破坏,破坏面从扩大头底面边缘向斜下方扩展,在水平方向影响范围达到最大后逐渐向桩内侧收缩;荷载作用越大,地基破坏区域越大,相应的极限抗压承载力也越大;持力层厚度增加,扩大头分担的荷载比例增大,分担的荷载达到稳定需要的桩顶位移也越大,H_h=0.5 D试验扩大头分担的荷载比例稳定时为60%,对应的桩顶位移约为29 mm;桩顶位移达到33 mm后,H_h=1.0~3.0 D试验稳定在63%~65%之间;通过细观颗粒流理论对砂土移动特性的研究发现,持力层厚度从0.5 D增大至2.0 D,破坏面的起始扩展角度从31°增大至42°。数值模拟研究结果与模型试验数据吻合效果良好,证明该方法分析多层砂土地基扩底桩单桩抗压荷载传递机理是可行的。  相似文献   

微塑料对海洋桡足类摄食、排泄及生殖的影响     

于娟  许瑞  魏逾杰  田继远  王雪丹  来敬广  陈容  杨桂朋  张正雨  邵云粹 《中国海洋大学学报(自然科学版)》2020,50(3):73-80

微塑料污染目前成为海洋污染普遍关注的一个研究热点。本文在实验室内将青岛近海常见的海洋桡足类猛水蚤暴露于不同浓度的微塑料尼龙6中,研究了猛水蚤的摄食、排泄以及生殖的变化。研究结果表明,微塑料尼龙6对猛水蚤的摄食、排泄、生殖均产生不利的影响,并且存在剂量-效应关系。微塑料尼龙6对猛水蚤摄食率、滤水率、排粪率的24 h·EC 50分别为67.7、62.2、84.1 mg·L^-1,对猛水蚤抱卵率的144 h·EC 50为30.3 mg·L^-1。“饱食感”造成猛水蚤摄食率降低,从而能量和营养摄入不足可能是导致猛水蚤抱卵率降低的原因。猛水蚤对微塑料的摄食,导致猛水蚤排泄的粪便颗粒小型化,由长椭球体变为短小椭球体,可能与其粘度或物理结构的改变有关。暴露于尼龙6的猛水蚤的粪便体积和沉降速率显著低于未暴露微塑料的对照组。本实验结果对于研究微塑料对海洋桡足类以及滤食性浮游动物的生态毒理影响具有一定的帮助。  相似文献   

河北滦县－滦南县段滦河冲积平原水文地质特征分析     

杨庆洪  左文喆  祝彬 《地质与资源》2020,29(4):374

滦县－滦南县段滦河冲积平原铁矿群司家营铁矿区、马城铁矿区、坎上铁矿区、常裕铁矿、高官营铁矿总储量达几十亿吨，均覆盖有巨厚的第四系含水体，水文地质条件复杂，一直未能开采，成为“大水呆滞金属矿”，被视为开发过程中可能对区域水文地质环境造成较大影响的大水矿区，列为限制开采区，至本世纪初才相继解限投产.通过对司家营铁矿北区、司家营铁矿南区、马城铁矿区、坎上铁矿区、常裕铁矿、高官营铁矿等6处矿山的水文地质条件的分析，论述了滦县－滦南县段滦河冲积平原的水文地质特征.  相似文献   

黑龙江省兴凯湖平原土壤pH值时空变化特征分析     

杨泽  刘国栋  戴慧敏  张一鹤  许江  肖红叶 《地质与资源》2020,29(6):614-620

土壤酸碱度是影响土壤化学和生物学性质的一个重要因素，会对土地生产力产生重要影响.为掌握兴凯湖平原土壤pH值近年来变化状况，采用空间统计方法，基于全国第二次土壤普查及2019年土壤pH值数据，对工作区近30年土壤pH值时空变化特征进行了分析.结果表明：兴凯湖平原土壤pH值总体呈下降趋势，但整体上土壤pH值变幅不大，主要集中在±0.5～±1区间.从土地利用类型上来看，兴凯湖平原耕地区土壤pH值均呈下降趋势，西部的林地区pH值整体保持不变或略有上升.在兴凯湖平原4种主要土壤类型中，除暗棕壤外，其他3类土壤在近30年间pH平均值均呈下降趋势.因此需要对工作区土壤pH值变化的原因进行多因素、定量分析，以进一步明确具体原因.  相似文献   

A motion planning algorithm for the feed support system of FAST     

姚蕊  姜鹏  孙京海  于东俊  孙纯 《天文和天体物理学研究(英文版)》2020,(5):117-126

The paper relates to a motion planning algorithm for the feed support system of the Five-hundred-meter Aperture Spherical radio Telescope(FAST).To enhance the stability of the feed support system,the start/termination planning segments are adopted with an acceleration and deceleration section.The source switching planning adopts a combination of a line segment and focal segment to realize stable control of the feed support system.Besides,during the observation trajectory,a transition segment which is not used for observation data is planned with a required time.Through an example simulation,a smooth change is realized via the motion planning algorithm and presented in this paper.  相似文献   

西昌市泸山地区典型火后泥石流特征与成因机制研究——以响水沟左岸3^#支沟为例     

铁永波  徐如阁  刘洪  王家柱  巴仁基  向安平  杨剑  欧阳渊  张玙  王宏 《中国地质调查》2020,7(3):82-88

2020年3月30日,四川省凉山州西昌市经久乡发生森林大火,山火燃烧后形成的山火灰和烧焦土体在强降雨作用下极易启动形成泥石流,给西昌市泸山、琼海景区生态环境及当地百姓生命财产安全带来潜在地质灾害风险。以西昌市安哈镇响水沟左岸3^#支沟为例,采用野外调查、无人机航拍、地球物理探测、现场及室内实验等手段,对2020年5月1日该沟发生的火后泥石流流体特征、动力学参数及成因机制等开展研究,系统分析了山火及降雨过程对该次火后泥石流形成的影响机制。该成果可为系统认识西昌泸山“3·30”山火地区火后泥石流的运动特征、物源规模、激发雨量及其发展趋势研判提供科学参考。  相似文献   

Data‐driven taxonomy matching of asteroid and meteorite     

Yuki Saito  Peng K. Hong  Takafumi Niihara  Hideaki Miyamoto  Kenji Fukumizu 《Meteoritics & planetary science》2020,55(1):193-206

The matching of asteroids and meteorites is a significant step toward a better understanding of the origin, structure, and history of the solar system. We propose a data‐driven approach for investigating common taxonomic structure between asteroids and meteorites; C‐, S‐, and V‐type for the former, and carbonaceous chondrite, ordinary chondrite, and howardite‐eucrite‐diogenite (HED) meteorite for the latter. In the numerical experiments, by checking whether the taxonomy information of meteorites improves classification for asteroid data, we examine the existence of common structure over the two domains. For this purpose, we compare the resultant accuracies of two clustering methods which are with/without the guidance of meteorite data. We observe that the guidance of meteorite taxonomy improves the accuracy for classifying asteroids, either with the reflectance spectra or major chemical compositions of meteorites. This fact serves as a piece of evidence that there is a common taxonomic structure and links between meteorites and asteroids, supporting a long‐standing hypothesis.  相似文献