基于负样本多通道优化SSD网络的钢铁厂提取     

卢凯旋  李国清  陈正超  昝露洋  李柏鹏  高建威 《中国科学院大学学报》2020,37(3):352-359

准确提取钢铁厂对去产能监测和环境保护具有重要意义。传统的人工目视解译方法效率低、成本高，无法满足开展大区域钢铁厂监测的需求。以深度学习目标检测网络SSD为基础，构建面向遥感影像钢铁厂提取的深度学习目标检测网络，提出maxout模块，将负样本通路优化为多分支结构，突出难分负样本特征并提升网络对无用特征的抵制效果。利用国产GF-1数据对京津冀地区的钢铁厂进行快速自动提取实验。与人工解译的钢铁厂点位数据的对比表明，该目标检测方法的提取精度达到80%以上。  相似文献   

一种在轨海上多运动舰船目标检测和跟踪方法     

矫腾章  胡玉新  吕鹏  张凯  台宪青 《中国科学院大学学报》2020,37(3):368-378

传统遥感卫星图像处理采用在地面进行目标检测和跟踪处理的模式，卫星将拍摄到的图像数据下传至地面数据处理中心，由地面数据处理系统对接收到的遥感图像数据进行目标检测和跟踪。然而，随着遥感图像分辨率的不断提高，需要下传的数据量增大，导致星地数据传输的时间大大增加，最终造成目标检测和跟踪的时效性降低。针对上述问题，提出一种基于多特征量判别的Canny边缘检测和联合概率数据关联的在轨海上多运动舰船目标检测和跟踪方法。将该方法利用中国科学院微小卫星创新研究院的高分微纳卫星实测数据在模拟星载的嵌入式开发平台上进行验证，结果表明该方法能够在轨对海上多运动舰船目标进行快速、准确的检测和跟踪。  相似文献   

The application of inertial measurement units and functional principal component analysis to evaluate movement in the forward 3½ pike somersault springboard dive     

Cherie Walker  John Warmenhoven  Peter J. Sinclair 《Sports biomechanics / International Society of Biomechanics in Sports》2019,18(2):146-162

Based on technological and analytical advances, the capability to more accurately and finitely examine biomechanical and skill characteristics of movement has improved. The purpose of this study was to use Inertial Measurement Units (IMUs) and Functional Principal Components Analysis (fPCA) to examine the role of movement variability (assessed via angular velocity), on 2 divers (1 international level; 1 national) performing the forward 3½ pike somersault dive. Analysis of angular velocity curves during ive-flight identified 5 fPCs, accounting for 96.5% of movement variability. The national diver’s scatter plots and standard deviation of fPC scores illustrated larger magnitudes of angular velocity variability across dive flight. For fPC1 and fPC3, magnitudes of SD variability were 282.6 and 201.5, respectively. The international diver illustrated more consistent angular velocity profiles, with clustering of fPCs scores (e.g., fPC1 & 3 = SD’s of 75.2 & 68.0). To account for lower variability in the international diver, the ability to better coordinate movement sequences and functionally utilise feedback in response to initiation of the somersault position is highlighted. Overall, findings highlight how both IMUs and fPCA can more holistically and finitely examine the biomechanical and skill characteristics of movement sequences with the capability to inform athlete development.  相似文献   

研究数据使用统计新标准及其应用案例研究     

林伟明  叶兰 《图书情报工作》2019,63(16):32-42

[目的/意义]介绍Make Data Count与COUNTER联合推出的《研究数据使用统计实施规范》，为数据级别计量提供新指标与新视角。[方法/过程]通过对标准文本的分析，介绍该规范的提出背景、目标、范围、相关概念及核心内容，通过案例剖析Dash、DataONE、Zenodo及其他7个数据存储库对《规范》的应用情况。[结果/结论]研究数据的使用统计具有其独特之处，《规范》的推出可对数据引用及数据替代计量形成补充进而描述完整的科研学术影响力。目前遵循该规范的数据存储库还不多，为推动数据使用计量的应用，需要标准组织、科研人员、机构库及数据存储库、出版商、科研机构及资助机构、图书馆等不同利益相关者在数据产生、管理、传播与利用等环节的相互合作。  相似文献   

我国图书情报领域期刊论文的科学数据引用特征研究     

丁文姚  李健  韩毅 《图书情报工作》2019,63(22):118-128

[目的/意义] 探索期刊论文科学数据引用特征与规律不仅有助于描述学科领域对科学数据的利用情况，还能够揭示学术成果表达中的数据引用模式。[方法/过程] 以我国图书情报领域6种期刊2017年与2018年第一期刊载论文为样本，结合国家标准《信息技术科学数据引用》的引用元素，采用内容分析法从9个维度对样本论文的科学数据引用行为进行数据编码，应用统计学方法描述图书情报领域期刊论文科学数据引用特征并探索不同维度特征间的关联关系。[结果/结论] 图书情报领域期刊论文广泛引用来自国内外的统计整理类科学数据，对期刊论文中个人研究科学数据的引用量较大；科学数据引用标注方式与科学数据类型存在一定对应关系，但多样化的标注方式缺乏统一性；二手引用现象较为突出，二手引用程度与科学数据创建者类型相关。  相似文献   

基于用户兴趣度量的知识发现服务精准推荐     

丁梦晓  毕强  许鹏程  李洁  牟冬梅 《图书情报工作》2019,63(3):21-29

[目的/意义]针对当前知识发现服务中存在的个性化程度不高和推荐效果不佳等问题，提出一种基于用户兴趣度量和内容分析的推荐算法。[方法/过程]文章通过特征词分布、LDA主题分布、引文结构网络三个维度构建学术资源模型，并通过对用户行为的度量，计算用户对其浏览学术资源的兴趣度，结合学术资源模型构建用户兴趣模型。将用户兴趣模型与学术资源模型匹配，计算其相似度，得到用户对每条学术资源的兴趣值，最后将兴趣值最高的TOP-N学术资源推荐给用户。[结果/结论]通过实验检验算法的有效性和推荐准确率，结果显示，本文从实时动态度量兴趣的角度，提出的推荐算法能较好地预测用户兴趣，推荐效果显著，为实现发现服务精准推荐提供思路。  相似文献   

Writer identification approach based on bag of words with OBI features     

Amal Durou  Ibrahim Aref  Somaya Al-Maadeed  Ahmed Bouridane  Elhadj Benkhelifa 《Information processing & management》2019,56(2):354-366

Handwriter identification aims to simplify the task of forensic experts by providing them with semi-automated tools in order to enable them to narrow down the search to determine the final identification of an unknown handwritten sample. An identification algorithm aims to produce a list of predicted writers of the unknown handwritten sample ranked in terms of confidence measure metrics for use by the forensic expert will make the final decision.Most existing handwriter identification systems use either statistical or model-based approaches. To further improve the performances this paper proposes to deploy a combination of both approaches using Oriented Basic Image features and the concept of graphemes codebook. To reduce the resulting high dimensionality of the feature vector a Kernel Principal Component Analysis has been used. To gauge the effectiveness of the proposed method a performance analysis, using IAM dataset for English handwriting and ICFHR 2012 dataset for Arabic handwriting, has been carried out. The results obtained achieved an accuracy of 96% thus demonstrating its superiority when compared against similar techniques.  相似文献   

Co-occurrence of adverse childhood experiences and its association with family characteristics. A latent class analysis with Dutch population data     

《Child abuse & neglect》2019

BackgroundAlthough adverse childhood experiences (ACEs) are relatively common among children, there is limited knowledge on the co-occurrence of such experiences.ObjectiveThe current study therefore investigates co-occurrence of childhood adversity in the Netherlands and whether specific clusters are more common among certain types of families.Participants and SettingRepresentative data from the Family Survey Dutch population 2018 (N = 3,128) are employed.MethodWe estimate Latent Class Analysis (LCA) models to investigate co-occurrence of ACEs. As ACEs we examine maltreatment, household dysfunction, demographic family events, as well as financial and chronic health problems. Gradual measures for maltreatment and financial problems are studied to make it possible to differentiate with regard to the severity of experiences.ResultsOur results show that four ACE clusters may be identified: ‘Low ACE’, ‘Moderate ACE: Household dysfunction’, ‘Moderate ACE: Maltreatment’ and ‘High ACE’. Regression analyses indicated that mother’s age at first childbirth and the number of siblings were related to experiencing childhood adversity. We found limited evidence for ACEs to be related to a family’s socioeconomic position.ConclusionThe found clusters of ACEs reflect severity of childhood adversity, but also the types of adversity a child experienced. For screening and prevention of childhood adversity as well as research on its consequences, it is relevant to acknowledge this co-occurrence of types and severity of adversity.  相似文献   

Location extraction from tweets     

Thi Bich Ngoc Hoang  Josiane Mothe 《Information processing & management》2018,54(2):129-144

Five hundred million tweets are posted daily, making Twitter a major social media platform from which topical information on events can be extracted. These events are represented by three main dimensions: time, location and entity-related information. The focus of this paper is location, which is an essential dimension for geo-spatial applications, either when helping rescue operations during a disaster or when used for contextual recommendations. While the first type of application needs high recall, the second is more precision-oriented. This paper studies the recall/precision trade-off, combining different methods to extract locations. In the context of short posts, applying tools that have been developed for natural language is not sufficient given the nature of tweets which are generally too short to be linguistically correct. Also bearing in mind the high number of posts that need to be handled, we hypothesize that predicting whether a post contains a location or not could make the location extractors more focused and thus more effective. We introduce a model to predict whether a tweet contains a location or not and show that location prediction is a useful pre-processing step for location extraction. We define a number of new tweet features and we conduct an intensive evaluation. Our findings are that (1) combining existing location extraction tools is effective for precision-oriented or recall-oriented results, (2) enriching tweet representation is effective for predicting whether a tweet contains a location or not, (3) words appearing in a geography gazetteer and the occurrence of a preposition just before a proper noun are the two most important features for predicting the occurrence of a location in tweets, and (4) the accuracy of location extraction improves when it is possible to predict that there is a location in a tweet.  相似文献   

From user-generated data to data-driven innovation: A research agenda to understand user privacy in digital markets     

《International Journal of Information Management》2021

In recent years, strategies focused on data-driven innovation (DDI) have led to the emergence and development of new products and business models in the digital market. However, these advances have given rise to the development of sophisticated strategies for data management, predicting user behavior, or analyzing their actions. Accordingly, the large-scale analysis of user-generated data (UGD) has led to the emergence of user privacy concerns about how companies manage user data. Although there are some studies on data security, privacy protection, and data-driven strategies, a systematic review on the subject that would focus on both UGD and DDI as main concepts is lacking. Therefore, the present study aims to provide a comprehensive understanding of the main challenges related to user privacy that affect DDI. The methodology used in the present study unfolds in the following three phases; (i) a systematic literature review (SLR); (ii) in-depth interviews framed in the perspectives of UGD and DDI on user privacy concerns, and finally, (iii) topic-modeling using a Latent Dirichlet allocation (LDA) model to extract insights related to the object of study. Based on the results, we identify 14 topics related to the study of DDI and UGD strategies. In addition, 14 future research questions and 7 research propositions are presented that should be consider for the study of UGD, DDI and user privacy in digital markets. The paper concludes with an important discussion regarding the role of user privacy in DDI in digital markets.  相似文献