基于用户自然标注的TF-IDF辅助标引算法及实证研究

[目的/意义] 从用户角度出发,研究基于用户自然标注的TF-IDF辅助标引算法。[方法/过程] 首先以核心期刊论文中作者标注的关键词和分类号为源数据,通过对关键词词频进行统计,使用TF-IDF算法构建用户标注词表、形成标引知识库,然后通过IK Analyzer分词软件对待标引的科技项目数据进行切词和停用词处理,进而使用TF-IDF算法和位置加权算法提取科技项目数据的特征词,最终实现对科技项目数据进行关键词和分类的同步标引。[结果/结论] 实验结果表明,机标关键词与人标关键词的相似比在60%以上的科技项目数据占总数的68.1%,机标分类号与人标分类号前三位一致的占总数的83.9%,结果表明基于用户自然标注数据并采用TF-IDF算法在关键词和分类标引方面是可行的。     

面向学术图书的Altmetrics指标分析

[目的/意义] 探讨Altmetrics指标对学术图书影响力进行评价的有效性,为学术图书评价工作提出合理建议。[方法/过程] 获取Twitter提及量、Mendeley阅读量、在线书评数量以及馆藏量指标数据,对数据集的覆盖率、分位数等统计量分析后,将被引频数与Altmetrics指标进行了指标间相关系数检验,再对高Altmetrics指标值的学术图书进行年份分布、学科差异及图书主题等实证分析,探究各指标在学术图书影响力评价中的应用。[结果/结论] 传统计量指标被引频数与Altmetrics指标之间的相关性较低,说明Altmetrics可以作为学术图书评价的一个新视角,不同Altmetrics指标反映了学术图书影响力的不同维度。未来的学术图书影响力评价建议结合学术图书的年份、学科等特征,将传统的引文与Altmetrics指标相结合,探索更全面有效的评价机制。     

高校图书馆社会化服务中用户信用体系研究

[目的/意义]针对我国高校图书馆向校外用户提供借阅服务时收取押金的现象,寻求可覆盖大部分用户的免押金信用借阅体系。[方法/过程]分析高校图书馆所遇信用问题的产生原因,对比国内外高校图书馆面向校外读者的借阅制度。[结果/结论]在信息化社会中,高校图书馆社会化服务中的用户信用体系由用户信用信息体系、用户信用评价体系、用户失信惩罚制度、用户信用教育体系、信用风险转嫁机制等构成。     

公共图书馆转型背景下的噪音问题——基于微博的调查

文章以用户在微博上发表的对公共图书馆噪音的看法为数据来源,使用内容分析和情感分析方法研究用户对噪音的认知,分析噪音的来源。研究发现:用户对公共图书馆的噪音问题表示关注和不满;大城市公共图书馆被用户提及最多;用户认为安静才是公共图书馆的应有之义;在用户认为的噪音来源中,用户因素远高于其他因素,最主要的噪音源是少年儿童吵闹。对策包括:转变用户对公共图书馆严肃安静的刻板印象;依据用户需求合理规划空间;改进馆内管理工作。     

生存分析法在学术期刊评价上的应用初探

期刊评价中文献老化理论视角的研究相对薄弱。本文探究生存分析法在学术期刊评价上的应用,在理论上 克服了因领域差异导致的壁垒,并选取CSTPCD中30家中文物理学科期刊作为样本数据集进行实证研究。通过对论文 被引时序数据进行分析,研究物理学科文献老化规律,并选取样本期刊进行对比分析,以探究期刊评价的新思路。     

Technological research in the EU is less efficient than the world average. EU research policy risks Europeans’ future

We have studied the efficiency of research in the EU by a percentile-based citation approach that analyzes the distribution of country papers among the world papers. Going up in the citation scale, the frequency of papers from efficient countries increases while the frequency from inefficient countries decreases. In the percentile-based approach, this trend, which is uniform at any citation level, is measured by the e_p index that equals the P_{top 1%}/P_{top 10%} ratio. By using the e_p index we demonstrate that EU research on fast-evolving technological topics is less efficient than the world average and that the EU is far from being able to compete with the most advanced countries. The e_p index also shows that the USA is well ahead of the EU in both fast- and slow-evolving technologies, which suggests that the advantage of the USA over the EU in innovation is due to low research efficiency in the EU. In accord with some previous studies, our results show that the European Commission’s ongoing claims about the excellence of EU research are based on a wrong diagnosis. The EU must focus its research policy on the improvement of its inefficient research. Otherwise, the future of Europeans is at risk.     

Exploring policies for the reduction of child physical abuse and neglect

Policies can be powerful tools for prevention given their potential to affect conditions that can improve population-level health. Given the dearth of empirical research on policies’ impacts on child maltreatment, this article (a) identifies 37 state policies that might have impacts on the social determinants of child maltreatment; (b) identifies available data sources documenting the implementation of 31 policies; and (c) utilizes the available data to explore effects of 11 policies (selected because they had little missing data) on child maltreatment rates. These include two policies aimed at reducing poverty, two temporary assistance to needy families policies, two policies aimed at increasing access to child care, three policies aimed at increasing access to high quality pre-K, and three policies aimed at increasing access to health care. Multi-level regression analyses between within-state trends of child maltreatment investigation rates and these 11 policies, controlling for states’ childhood poverty, adults without a high school diploma, unemployment, child burden, and race/ethnicity, identified two that were significantly associated with decreased child maltreatment rates: lack of waitlists to access subsidized child care and policies that facilitate continuity of child health care. These findings are correlational and are limited by the quality and availability of the data. Future research might focus on a reduced number of states that have good quality administrative data or population-based survey data on child maltreatment or reasonable proxies for child maltreatment and where data on the actual implementation of specific policies of interest can be documented.     

基于Mystery Shopping的公共图书馆服务质量评估

解析Mystery Shopping评估检测法,研究威尔士公共图书馆服务质量评估Mystery Shopping项目,揭示该项目对国内公共图书馆服务质量评估的启示。     

职业训练的效益评估——以企业内训为例

近年来,职业训练训后评估逐渐被重视,例如美国训练与发展协会(ASTD)及台湾人才发展质量管理系统(TTQS)等,皆关注参训者于训练时所习得的新知识或技术是否能应用于职场中,以利直接提升企业的实质效益。本研究于2014年5⁶月份期间,协助台湾地区65家事业单位进行训后追踪评估,总共搜集了65位企业主管及429位参训学员评估训练成效。研究发现,学员认为参训后在新知识及技术的获得、受训内容应用于实务中、及提升工作表现等方面,具有最明显的改变;此结果与企业主管评估参训学员训后能力提升的结果相符,主管认为员工参训后在训练内容的应用、获得新知识与技术、提升工作表现、及问题解决能力等皆有不错表现。另外,企业主管也认为训练除了能提升员工表现以外,在组织目标达成、组织营运绩效、及组织营运成本等,皆有正向的影响及改善。整体上,参训学员及企业主管对于训练计划成果持正向支持的态度,并乐于继续参与训练课程以加强组织竞争力。     

基于粗糙集的决策规则可靠性评估方法

粗糙集理论是处理不精确、不准确数据的有效方法，但是通过粗糙集方法获得的决策规则对于不完整的信息承统和随机数据也是不确定的。本文描述了一个用于决策规则的可靠性评价方法。该方法独立于任何专用引申规则方法，并且参数能够被调整以适应不同的信息系统。