共查询到20条相似文献，搜索用时 246 毫秒
1.
[目的/意义]国家政府、大中型企业以及研究机构面对技术难题,如何找到合适的专家是迫切需要解决的问题。面对需要运用多学科知识来解决的综合性复杂难题,寻找到多专长专家显得尤为重要,寻找合适的方法识别出多专长专家是本研究的目的。[方法/过程]利用专家所发表的学术论文数据,通过抽取专家有代表性的研究专长特征,基于TFIDF加权的重叠K-means聚类算法对专家进行重叠聚类划分,挖掘出专家的多个研究专长,进而识别出多专长专家。[结果/结论]研究结果表明TFIDF加权的重叠K-means聚类算法在查准率、召回率和F值上有良好的表现,可以识别多专长专家。 相似文献
2.
聚类有效性指标用于评价聚类质量和确定最佳聚类数,针对包含大小和密度差异性较大数据类的数据集,在分析了传统模糊聚类有效性指标不足的基础上,提出了一个同时考虑紧致性、重叠度和分离性的聚类有效性指标COS.类内紧致性用一定阈值内的隶属度之和与最大类内距离之比表示,一定阈值内各样本同属于两个类的隶属度差异反映了这两个类的重叠度,类间分离性的度量为最小类间距离,使COS指标值最大的聚类数即为最佳聚类数.在四个人工数据集和iris真实数据集上利用模糊C均值算法进行聚类实验的结果表明,COS指标可以有效发现小类和低密度类. 相似文献
3.
组织专家的检索系统设计与实现 总被引：3，自引：0，他引：3
组织专家的检索是当前垂直信息检索研究的热门领域,也是组织信息检索研究的重要内容。本文总结了目前国内外组织专家的检索研究现状,分析了构建组织专家的检索系统的需求和挑战,采用基于相关文档集的归并排序法,利用组织内外的网页和期刊数据库等信息智能识别专家的专长,并基于此设计了从数据资源采集、规整、索引、检索到可视化等整个过程的组织专家的检索系统模型及以武汉大学为例的检索系统平台。 相似文献
4.
[目的/意义]探索领域知识发展过程中的聚类演化问题有助于揭示知识聚类的特征和规律,对于掌握知识生长演进过程中关联知识的聚集具有重要意义。[方法/过程]以复杂网络的思想为基础,基于标签邻接关系的发生值构建时间序列领域知识网络。即依据网络模体的理论,采用网络聚类系数的分析方法,对领域知识网络进行动态跟踪与分析;结合网络密度、特征路径长度、节点度值、封闭三元组等指标,从随机因素、度相关性、邻近关联3个方面对领域知识发展过程中的聚类演化现象进行分析。[结果/结论]研究结果表明：①领域知识在发展进程中始终保持较高的聚类性;②领域知识的聚类性同时包含随机性与结构性（非随机性）两方面因素; ③领域知识聚类的动态状态在小世界网络和无标度网络之间摇摆演化; ④领域知识的聚类状态在网络全局和局部节点之间表现出一定的差异性。 相似文献
5.
利用聚类分析工具可以从海量的、日益增长的基因表达数据中解析出其中可能的编码基因及其生理功能,这是生物信息学中一项很有意义的工作。模糊聚类算法是一种常用的基因表达数据聚类分析工具,可以发现重叠的基因簇,它不强制将每个基因归入某个具体的聚类中,而是计算每个基因对各个类的隶属度。本文分析了基因表达数据的模糊聚类分析方法及其重要应用。 相似文献
6.
对BWP方法进行研究,从嵌入式NSTL个性化推荐的文本聚类需求入手,分析BWP方法的不足,提出一种改进的K-means算法最佳聚类数确定方法。对单一样本类的类内距离计算方法进行优化,扩展BWP方法适用的聚类数范围,使原有局部最优的聚类数优化为全局最优。实验结果可以验证该方法具有良好性能。 相似文献
7.
8.
针对热点主题发现是在聚类算法的基础上实现的特点,将改进后的蚁群聚类算法引入到该研究中,同时提出类别关注度（CAD）的概念,以此来判定类别的热门程度并区分出热门类别和冷门类别,在此基础上抽取热点主题集。实验结果表明改进后的蚁群聚类算法对热点主题的发现有一定的效果,对其他仿生优化聚类算法的引入有借鉴意义。 相似文献
9.
10.
11.
This article explores the way librarians define, leverage, and amplify expertise in a twenty-first century academic library. An expert team comprised of a nursing librarian, online learning librarian, information-literacy librarian, and assessment librarian sorted the learning outcomes from the Information-Literacy Competency Standards for Nursing created by the Health Sciences Interest Group taskforce of the Association of College and Research Libraries (ACRL) by grade-levels. Results found distinguishing experts within a library supports the customization of scaffolded instruction. Additionally, using expert teams in academic libraries supports the larger mission of universities to integrate libraries into teaching and research. 相似文献
12.
[目的/意义]提出一种基于专家个体多维特征刻画的电子政务项目评审专家组推荐方法,提升专家组间项目评审的一致性水平。[方法/过程]以专家个体的长期评审意见为数据源,利用意见挖掘技术实现知识元识别与情感极性获取;构造专家的领域知识结构并动态迭代更新;利用统计分析刻画专家知识水平、评审深刻性、情感风格、领域专长特征,实现基于科学计量的专家特征刻画并以此为基础进行专家组合的推荐。[结果/结论]本文的方法注重专家组的多维特征均衡,对电子政务项目评审具有很好的问题针对性,并在实践中取得了良好的应用效果。 相似文献
13.
[目的/意义] 为投稿论文遴选出合适的审稿专家是论文发表过程中关键的一环。随着投稿论文和候选评审专家数量的持续增长,人工指定评审专家的方法在准确性和公平性上的弊端日益显露出来。因此,为进一步提高专家评审的客观性和准确性,笔者从专家知识与专家权威度两个维度对专家建模,并以此为依据为不同主题的投稿论文遴选推荐评审专家。[方法/过程] 首先分析专家知识以及投稿论文的研究内容,并提取两者涉及的多个子研究主题;然后,计算专家知识对投稿论文子主题的覆盖度,并提出融合主题特征与时间特征的权威度算法TTAM来分析专家权威度;最后,提出融合主题覆盖度和专家权威度的专家推荐框架CAUFER,综合考虑覆盖度和权威度两个因素为投稿论文推荐合适的评审专家。[结果/结论] 实验结果表明,与经典的基于向量空间模型、语言模型和作者主题模型3种专家推荐算法相比,笔者提出的算法能够较好地提高专家与投稿论文的匹配度,并可据此追踪专家权威度的变化,刻画专家在特定主题下的权威度,进一步提高专家推荐的准确性和科学性。 相似文献
14.
[目的/意义]针对科研项目评审专家选取工作中的回避问题,试图提出一套可行的用于确定回避专家的模型方法,避免人际关系对评审工作的干扰,保障科研评价的规范性和公正性。[方法/过程]首先以专家与申请人之间的论文合作关系为切入点,以合作强度、合作作者深度、方向匹配度和评审业绩为主要指标;然后采用综合权重和修正TOPSIS法对各指标进行分析,构建面向科研项目的专家回避量化模型;最后根据规定的专家组成员数量,考虑年龄、机构、性别等分布情况,明确应该回避的专家。[结果/结论]经检验,该模型在给出回避专家名单时,可以有效避免"唯合作论"的问题,回避理由充分,符合科技管理部的专家选取要求,能够为专家遴选工作提供一定的决策支撑。 相似文献
15.
16.
Elizabeth M. Johns Antoinette S. Ungaretti 《Journal of Library & Information Services in Distance Learning》2019,13(1-2):21-39
Research librarians often have a specialty in certain discipline, and many academic libraries thrive on this type of arrangement to support their researchers. But more often, new interdisciplinary programs are developed that require the expertise of more than one subject expert for support. In particular, students in a Master of Education in the Health Professions (MEHP) program have unique, complicated research challenges that can only be addressed with careful collaboration amongst multiple subject experts. However, this program and the library saw the need for additional expertise to support their students, and began a collaboration with the medical library, bringing a clinical informationist onto the library support team. Together, the education librarian, clinical informationist, and program faculty collaborate to provide excellent services to help students in their research throughout the program. 相似文献
17.
In many realistic settings of expert finding, the evidence for expertise often comes from heterogeneous knowledge sources.
As some sources tend to be more reliable and indicative than the others, different information sources need to receive different
weights to reflect their degrees of importance. However, most previous studies in expert finding did not differentiate data
sources, which may lead to unsatisfactory performance in the settings where the heterogeneity of data sources is present.
In this paper, we investigate how to merge and weight heterogeneous knowledge sources in the context of expert finding. A
relevance-based supervised learning framework is presented to learn the combination weights from training data. Beyond just
learning a fixed combination strategy for all the queries and experts, we propose a series of discriminative probabilistic
models which have increasing capability to associate the combination weights with specific experts and queries. In the last
(and also the most sophisticated) proposed model, the combination weights depend on both expert classes and query topics,
and these classes/topics are derived from expert and query features. Compared with expert and query independent combination
methods, the proposed combination strategy can better adjust to different types of experts and queries. In consequence, the
model yields much flexibility of combining data sources when dealing with a broad range of expertise areas and a large variation
in experts. To the best of our knowledge, this is the first work that designs discriminative learning models to rank experts.
Empirical studies on two real world faculty expertise testbeds demonstrate the effectiveness and robustness of the proposed
discriminative learning models. 相似文献
18.
19.
基于主题图构建知识专家学术社区研究 总被引：2，自引：1，他引：1
20.
Eileen Wood Domenica De Pasquale Julie Lynn Mueller Karin Archer Lucia Zivcakova Kathleen Walkey 《The Reference Librarian》2013,54(3):182-204
The relative contributions of expertise in search skills and domain knowledge were examined when using the Internet to find information. Four conditions were compared: expert searchers/high domain knowledge; expert searchers/low domain knowledge; novice searchers/high domain knowledge; and novice searchers/low domain knowledge. Search outcomes and verbal protocols were analyzed. The combination of search expertise and high domain knowledge yielded the most efficient searches. Higher search expertise yielded access to sites rated more accurate and credible. High domain knowledge yielded sites rated more thorough. Verbal protocols depicted searching as a complex decision process. Findings have implications for instructional support. 相似文献