创建学习型图书馆 增强图书馆竞争优势

针对目前网络环境下图书馆面临的竞争和挑战，提出了创建学习型图书馆，以增强图书馆的竞争优势的思想，并以国外学习型组织的理论为依据，论述了什么是学习型图书馆以及如何创建学习型图书馆。     

对高校图书馆将信息咨询服务推向市场的探讨

本文提出了开辟社会服务、提供主动服务、扩大智慧服务、发展综合服务、组织联合服务、引入有偿服务等高校图书馆信息服务的发展趋势,并对高校图书馆如何转变观念、建设内部管理和运行机制、发展具有高校图书馆特色的信息产业等问题作了探讨     

基于无监督图神经网络的学术文献表示学习研究

学术文献特征表示,是学术文献搜索、分类组织、个性化推荐等学术大数据服务的关键步骤。研究表明,图神经网络能够有效学习文献的特征表示,然而当前研究主要集中在有监督学习方法上,不仅对数据集的大小和质量的要求较高,且学习到的文献特征表示与具体任务高度耦合。基于此,本文将四种无监督图神经网络方法引入学术文献表示学习,从Cora、CiteSeer和DBLP (database systems and logic programming)数据集的引文网络、共被引网络和文献耦合网络中学习文献的表示向量,并应用于文献分类和论文推荐两大下游任务。研究结果表明,(1)深度互信息图神经网络适合于文献分类任务,对抗正则化变分图自编码器则在论文推荐任务上性能更佳;(2)Cora数据集上的结果表明,相较于共被引和文献耦合网络,引文网络更适合于学习通用的文献表示向量。     

融合深度学习和链路预测的细粒度技术预测研究——以合成生物技术为例

[目的/意义]面向专利文本进行更细粒度的技术实体识别和技术预测,利于更详细地把握专利技术布局与趋势。[方法/过程]首先利用深度学习方法自动识别专利技术术语类实体,通过实验对比多组深度学习算法的优劣。其次,提出新的半监督标注和自定义标注方案,提高人工标注效率。最后,执行训练得到的最优模型,结合链路预测方法,对合成生物技术进行细粒度的技术预测。[结果/结论]实证结果表明RoBERTa-BiLSTM-CRF模型更适用于语义复杂的专利技术实体识别,F1值可达到86.8%,技术识别结果比传统IPC分析方法更精细。同时,细粒度的技术预测结果表明,合成生物学的合成方法在不断改进创新,合成物研究向合成燃料发展。     

数据挖掘在生物医学工程文献检索中的应用

本文在概述数据挖掘技术的基本概念、分析方法的基础上,分析与比较了文本自动聚类算法;综述了国内外面向MEDLINE文献数据库的数据挖掘、知识提取研究;简述了可扩展标识语言(XML)的基本概念、文档格式及其在数据管理及数据挖掘中的应用.     

“数字图书馆”课程核心教学内容分析

论文以东南大学情报科学技术研究所硕士生学位课程"数字图书馆研究"教学内容为例,比较分析国内外数字图书馆课程核心教学内容的差别,希望我国能建立科学实用的课程体系,培养出符合实际需要的数字图书馆员.     

Two Thousand Students,One Librarian: Balancing Depth and Breadth of Library Instruction for Online Graduate Students

Many librarians struggle with how to balance scalable instructional offerings with more personalized instruction for online students—scale tends to come at the cost of personalization and vice versa. This paper will discuss three methods for connecting with online students that range from one-to-one personalization to widely scalable instruction: individual consultations, scaffolded instructional webinars, and asynchronous video tutorials. The student engagement with each type of instruction will be discussed, with ideas for future directions provided.     

Tools and Principles for Effective Online Library Instruction: Andragogy and Undergraduates

The library literature evaluating the efficacy of online tutorial methods and delivery tools is a valuable resource for librarians looking for information to inform their choices for developing asynchronous online instruction. This literature, however, examines a wide variety of delivery tools, methods, and student populations. As a result, it is difficult to draw definitive conclusions about any one tutorial tool or method. This article compares selected research and concludes that the consensus in the literature suggests that applying adult learning principles to library tutorials has a larger impact on effective online instruction than the individual tool.     

The Librarian BFF: A Case Study of a Cohort-Based Personal Librarian Program

This case study presents the development and implementation of a personal librarian program for a cohort-based, executive-style higher education administration doctoral program. Librarians and the program director collaborated to create a student-centered program based on individual research needs. The personal librarian program was designed to build a close relationship between the librarian and individual students; to help identify their research knowledge gaps; and to identify resources to meet each individual students' research needs.     

Does Physical Activity Enhance Learning Performance?: Learning Effectiveness of Game-based Experiential Learning for University Library Instruction

This paper reports research results of the proposed experiential learning using a game-based educational material. The educational material was an application running on a mobile device and was supported for learners to study university library instruction. During the experiential learning, the learners answered several quizzes provided from the educational material. The quizzes were including experiential contents such as actually to find an academic journal or to reserve a learning room. In order to evaluate learning effectiveness of the experiential learning, this research made a comparative experiment. The experimental group used the educational material and the control group used another educational material of e-learning (non-experiential) but the contents was almost same as the contents of the experiential learning. The comparative experiment took several evaluation elements such as pre-/post-test scores, delayed-test scores and learning motivation scores based on IMMS of ARCS model. The experimental results revealed that experiential learning was superior for the several elements of these metrics. This research, furthermore, analyzed learner's operation data to the educational material and learner's behavior data during the learning in the experimental group. The analyzing results showed three types of the learning behaviors and indicated that an instructor needs to adopt suitable instructional design to each type to bring superior learning effectiveness.