首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献,搜索用时 240 毫秒
1.
以大众分类(folksonomy)为基础的社会化网络中,通常用标签对资源特征和用户喜好进行描述,本文将标签语义引入传统的推荐系统,从用户的标注行为入手,提出一种基于超图投影的推荐方法,该方法利用投影图中节点的连边权重进行节点相似性度量,使待推荐对象在投影图上随机游走,根据待推荐对象在节点上停留概率获得推荐,实验结果表明标签的引入提高了推荐质量,算法在精确性和多样性上均有很好的改进.  相似文献   

2.
常江  徐帅 《青年记者》2017,(21):20-22
短视频新闻:在移动化情境下追求优质的用户体验 短视频新闻通常是指长度以秒计算,总时长一般不超过5分钟,利用智能终端进行美化、编辑,并可在多种社交平台上实时分享的一种新型视频新闻产品.短视频新闻在形态上往往融合语音、视频、文字、音乐等元素,表达形式直观、立体,满足互联网移动化、碎片化的表达情境.美国互联网流量监测机构Comscore于2016年7月8日发布的数据显示,96%的18-34岁的用户从移动设备(智能手机、笔记本电脑等)上浏览新闻,其中35%的用户只从手机上获取新闻,94%的35-54岁用户使用移动设备浏览新闻.  相似文献   

3.
文章以档案信息网络服务为视角,在重新汇总《数字档案馆系统测试办法》中关于网站的评分指标的基础上,采用随机多次访问的方式,调查了29个全国示范数字档案馆门户网站的建设和运行情况。经分析,发现全国示范数字档案馆门户网站在网站链接可使用性、网站的定位、网站内容运行和维护等方面存在问题,提出明确网站定位、深挖数字档案内容、改善网站可使用性、提升用户使用体验等优化建议,以期为我国数字档案馆从建设阶段进入应用阶段的可持续发展助力。  相似文献   

4.
韩勇  吴瑞 《图书情报工作》2006,50(3):98-100
针对Web日志中事务的类与类之间边界的不确定性,提出使用粗糙理论对Web事务中用户存取模式进行聚类的方法;给出任意两个用户存取模式之间的相似性度量的定义,并在聚类的过程中把具有相同上近似的模式归为一类,以便根据用户的行为建立可适应的Web服务器。  相似文献   

5.
基于属性值偏好矩阵的协同过滤推荐算法   总被引:7,自引:2,他引:5  
传统的协同过滤推荐算法面临用户评分数据稀疏性和冷启动问题的挑战.针对上述问题,提出了基于属性值偏好矩阵的协同过滤推荐算法,首先采用奇异值分解(SVD)对用户-项目评分矩阵降维得到目标用户的初始邻居用户集,生成新的用户-项目评分矩阵;然后将用户评分映射到相应的项目属性值上,生成每个用户的属性值偏好矩阵,并基于属性值偏好矩阵进行用户相似性度量,从而缓解了评分数据稀疏性;将新项目的属性值与用户的属性值偏好矩阵进行匹配,从而找出匹配度最高的前N个用户作为新项目的推荐受众.实验结果表明了该算法的有效性.  相似文献   

6.
大多数传统的数据挖掘算法,通常是在二进制值的事务数据库中在单一层次上发现属性之间的关联关系即关联规则,但是大多数的数据库包含有大量量化的值。通常人们采用分区的方法处理量化值,然而这种处理方法带来分区过硬的问题。本文使用模糊的方法从Web日志中发现模糊泛化的Web网页之间的关联规则,它们能够体现出带模糊浏览时间的网页之间的关联关系。通过实例分析表明,该算法在可承受的计算时间内可对Web日志中的用户存取模式进行有效的模糊泛化关联规则的提取。  相似文献   

7.
关联规则兴趣度度量方法的比较研究   总被引:1,自引:1,他引:1  
关联规则挖掘是数据挖掘中重要的研究课题,已有许多有效的实现算法。然而,这些算法找到的关联规则数目太多,用户无法对其进行分析。为了克服这个问题,出现了一些关联规则衡量标准来分析规则的有趣性,在本文里我们在给出的实例上比较分析了一些关联规则客观兴趣度度量指标,提出了使用关联规则客观兴趣度度量指标的一些建议。  相似文献   

8.
在电子商务中,协同推荐技术能够帮助用户发现感兴趣的东西.在协同推荐中,通常采用最近邻居的方法来产生推荐.随着商品数量的增多,协同推荐所需要的数据集也越来越稀疏,可用数据比例越来越少.为了解决这个问题,本文在传统的评分数据的基础上,引入用户的基本信息,对用户的基本信息进行离散化处理,将用户的基本信息转化成一个0、1的向量,在用户的信息的基础上计算最近邻居,根据最近邻居对用户缺失数据进行补充,在补充后的评分数据上进行聚类计算,并根据聚类结果对用户评分进行预测.实验表明引入用户的基本信息,并采用对基本信息离散化的处理方式进行缺失数据补充,在此基础上进行数据的聚类,能够提高预测评分的准确性.  相似文献   

9.
文章针对目前数字档案服务质量无法进行量化的问题,提出利用多层次分析法分析用户在使用数字档案过程中的用户体验评估模型.将无法定性的指标量化成数值,计算得出各评价指标在用户体验评估中的权重,通过一致性检验计算评估权重的可信度.以用户体验为中心,通过权重计算找出在使用电子档案资源过程中影响用户体验的主要因素,结合数字档案的使用说明计算过程,为数字档案资源的建设,提升数字档案资源的服务提供参考方法.  相似文献   

10.
陶剑文  潘红艳 《情报学报》2008,27(2):199-204
推荐系统是电子商务系统中最重要的技术之一.随着电子商务系统用户数目和商品数目日益增加,在整个商品空间上用户评分数据极端稀疏,传统的相似性度量方法均存在各自的弊端,导致推荐系统的推荐质量急剧下降.针对用户评分数据极端稀疏情况下传统相似性度量方法的不足,本文提出了一种基于相似项目与用户评分预测的协同过滤推荐算法,综合利用相似项目和相似用户评分信息预测用户对未评分项目的评分.通过聚类算法形成用户候选近邻集,减小了算法搜索空间,降低了最近邻用户的搜索时间,从而增强了算法的扩展性.实验结果表明,本算法可以有效解决用户评分数据极端稀疏情况下传统协同推荐算法存在的问题,显著提高推荐系统的推荐质量.  相似文献   

11.
12.
Discovery tools offer users a powerful way of searching library holdings as well as external databases and indexes. They are becoming an increasingly common part of the library user experience, and research on the usability of such tools is expanding. In 2012, a mid-sized academic research library implemented Primo Discovery and Delivery by Ex Libris and conducted a diagnostic usability test to investigate how the tool is used without instruction, to discover patterns in searching behavior, and to uncover how compatible Primo is with user search behavior. This article will describe the design and implementation of the usability study. Despite some design flaws, users adapt to the tool over time and can perform effective searches. This article also provides recommendations for future usability studies in specific targeted areas where problems were detected.  相似文献   

13.
ABSTRACT

Digital libraries and collections are a growing facet of today's traditional library. Digital library technologies have become increasingly more sophisticated in the effort to provide more and better access to the collections they contain. The evaluation of the usability of these technologies has not kept pace with technological developments, however, and the end-user has in some cases been left behind. This research study evaluates the usability of digital collections created using the CONTENTdm Digital Collection Management System, a software system used by the University of North Carolina at Chapel Hill. Specifically, this study addresses the following questions: Does CONTENTdm meet users' needs? Is the interface sufficiently intuitive for them to use it? Is the experience of exploring digital collections using CONTENTdm satisfying to users? Employing usability testing techniques with actual-end users, this study attempts to assess the efficacy of the CONTENTdm public interface as well as user attitudes toward it. Ten participants from three user groups—faculty, library science graduate students, and the general public—performed eleven tasks designed to test the key functions of CONTENTdm-created collections and then answered a series of questions about their experiences. Key findings from this study indicate that while the included digital collections are useful and desirable to end-users, the interface generated by CONTENTdm can be confusing even for those who have considerable experience using the Internet. Results from this study may be used to improve this software system and add to the literature surrounding the usability of digital libraries in general.  相似文献   

14.
Boopsie, Inc. develops mobile applications for libraries, universities, conferences, and businesses. The researchers in this study investigated the usability of the Boopsie mobile application for a medium-sized academic library. Based on data collected from a series of qualitative interviews with undergraduate students, this research captured the users' experiences finding several types of information using a mobile application created by Boopsie. Three thematic findings emerged from seven study participants: Libraries should carefully select resources and features for inclusion in mobile applications; the simplicity of the mobile application interface offers some advantages over a conventional library Web site; and users may find the application more compelling for in-depth research than for answering quick questions. As result of this research, one library made modifications to its Boopsie application to improve usability. The results offer librarians or libraries practical advice about the promises and pitfalls of implementing a mobile library application.  相似文献   

15.
Usability testing on library search tools was conducted with ten students and eighteen library staff members at Syracuse University. The study addressed three research questions: (1) Do the ways in which librarians carry out search tasks on the library Web site vary from those of student users?; (2) Are those variations indicative of different mental models, i.e. different experience with and knowledge about the content and search tools on the Web site?; and (3) If there are differences, what are the implications of those different models for the usability of the site by students? Participants were tasked by using the library's Web site to locate both known items and items about a subject in the library's collection. The two groups differed in the success with which they were able to complete the tasks, as well as the time used. There were differences in the selection of search tools, the syntax used within the search box, the expectations of how library resources are organized, and the level of persistence as tasks were performed. Students’ behavior is characteristic of different mental models for how search tools work. The results are applicable to the design of the library's Web site and have implications for how library staff conduct reference and instruction services.  相似文献   

16.
In this paper the authors present a preliminary discussion on some of the results from a survey aimed to explore, describe and explain some of the usability characteristics in digital libraries evaluation in the Mexican context. The study is framed in the evaluation of a multinational and monolingual digital library: the Miguel de Cervantes Virtual Library, from the University of Alicante in Spain. The evaluators were Mexican “expert” users (i.e. Spanish-speaking professional university librarians specialized in electronic reference services) who were asked to carry on an evaluation instrument based on usability criteria as taken from some models in developed countries. Some questions that might be answered with future research are outlined in this paper.  相似文献   

17.
主要探讨服务主导型数字图书馆的体系结构、服务功能和相关技术及问题等。十几年来,数字图书馆的研究和发展,将形成数字图书馆三种主流模式即特种馆藏型模式、服务主导型模式和商用文献型模式;服务主导型数字图书馆的体系结构:它以三种主要数字化资源为基础和由统一信息访问平台、网上参考咨询(VRS)平台为两翼组成;统一信息访问平台主要解决异构平台的信息资源检索,向用户提供方便检索的统一界面,提供不同数字图书馆的互操作。网上参考咨询平台主要解决用户在访问数字图书馆时的疑问。这两个平台对一个实用的数字图书馆是不可缺少的,在系统结构上有些是相互渗透的。文中还探讨数字图书馆建设中相关的10个问题。  相似文献   

18.
认为上世纪90年代中期我国的信息化发展战略、计算机和互联网的推广应用、信息技术与发达国家的同步发展、图书馆协作为数字图书馆建设提供成熟的政策环境条件,而促进社会经济和文化发展、提高国家软实力、提高社会发展过程中的信息服务满足能力、缓解图书馆建设资源投入中的不足和利用效率问题则成为建设数字图书馆的政策动因。同时,有针对性地探讨我国数字图书馆建设在发展普及、服务方式、管理机制等方面存在的一些问题。  相似文献   

19.
论文参照国际可用性标准,在充分考虑省级公共图书馆社会职能的基础上,结合信息社会图书馆用户的特点,从用户使用角度研究省级公共图书馆网站的可用性,并对广西壮族自治区图书馆网站进行可用性现状分析,提出其网站的可用性改进意见。  相似文献   

20.
上海交通大学数字图书馆雏型   总被引:14,自引:2,他引:14  
简要介绍上海交通大学数字图书馆雏型产生背景和概念模型, 它包括五个模块: 各种载体数字化, 大容量数据储存和管理, 数据访问, 数字化资料传递, 权限管理和版权保护; 上海交通大学数字图书馆雏型的体系结构包括网络拓扑结构、主干网的缚线系统和硬软件环境等; SJTUDL 的实用服务模式由数字化图书馆的特种馆藏, 商用网上联机电子出版物和Internet 文献信息资源三部分组成, 将用一个统一的界面向读者提供服务。  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号