Web内容挖掘在数字图书馆中的应用

Web信息是数字图书馆数字资源的重要来源，文章详细阐述了对Web内容中的文本信息的挖掘，包括：文本自动摘要、文本分类和文本聚类。在挖掘过程中基于用户需求和用户特征，针对文本分类，重点分析了分类过程和在数字图书馆中的应用；针对文本聚类，介绍两种基本聚类原理以及与文本分类的不同之处，重点论述其在数字图书馆中如何应用。最后提出内容挖掘与用户挖掘的结合更有助于服务用户。     

How do government agencies review and approve text content for publication on their Web sites? A framework to compare Web content management practices

This article describes a multicase field study of four public-facing Midwestern state agencies and how they organized the work of reviewing and approving textual content for publication on Internet Web sites. Drawing on field data and concepts from institutional theory, contingency theory, and organizational design, the paper generates a framework to document systematically the processes and practices involved in the review and approval of textual Web content. The framework is then used to compare textual content management practices across the four case sites and to begin to investigate the effects of variation in content management practices on characteristics of the available textual content. The article suggests how the framework could be used in future research to investigate how content practices affect content characteristics such as quality, quantity, and cost and to investigate how and why content management practices change over time.     

End user searching: A Web log analysis of NAVER,a Korean Web search engine

Transaction logs of NAVER, a major Korean Web search engine, were analyzed to track the information-seeking behavior of Korean Web users. These transaction logs include more than 40 million queries collected over 1 week. This study examines current transaction log analysis methodologies and proposes a method for log cleaning, session definition, and query classification. A term definition method which is necessary for Korean transaction log analysis is also discussed. The results of this study show that users behave in a simple way: they type in short queries with a few query terms, seldom use advanced features, and view few results' pages. Users also behave in a passive way: they seldom change search environments set by the system. It is of interest that users tend to change their queries totally rather than adding or deleting terms to modify the previous queries. The results of this study might contribute to the development of more efficient and effective Web search engines and services.     

Crafting the message: Controlling content on agency Web sites

While much research has focused on the new opportunities that government Web sites offer for greater citizen involvement and improved agency efficiency, less attention has been given to agency decisions about what to post on these Web sites. Here we use interviews with content managers in seven federal agencies to investigate the political and institutional influences behind decisions about Web content. We analyze the approval processes for new content and the emerging governance structures for evidence of greater centralization and political control or greater decentralization and autonomy for Web posters. In the end, it appears that institutional factors persist to influence content governance.     

A machine learning approach to sentiment analysis in multilingual Web texts

Sentiment analysis, also called opinion mining, is a form of information extraction from text of growing research and commercial interest. In this paper we present our machine learning experiments with regard to sentiment analysis in blog, review and forum texts found on the World Wide Web and written in English, Dutch and French. We train from a set of example sentences or statements that are manually annotated as positive, negative or neutral with regard to a certain entity. We are interested in the feelings that people express with regard to certain consumption products. We learn and evaluate several classification models that can be configured in a cascaded pipeline. We have to deal with several problems, being the noisy character of the input texts, the attribution of the sentiment to a particular entity and the small size of the training set. We succeed to identify positive, negative and neutral feelings to the entity under consideration with ca. 83% accuracy for English texts based on unigram features augmented with linguistic features. The accuracy results of processing the Dutch and French texts are ca. 70 and 68% respectively due to the larger variety of the linguistic expressions that more often diverge from standard language, thus demanding more training patterns. In addition, our experiments give us insights into the portability of the learned models across domains and languages. A substantial part of the article investigates the role of active learning techniques for reducing the number of examples to be manually annotated.     

A Collaborative Web

Abstract

The content of a law library's website is often developed as part of reference services. Overall design, development, and management of the website, however, may be better addressed independent of content. Overall website management may reside in different areas in different libraries. This article shares experiences at the University at Buffalo Law Library, where the placement of website management within technical services has allowed for increased collaboration across all departments.     

对网络环境下引文分析评价方法的再认识

引文分析评价方法目前已在许多领域中应用。文章认为,由于各种因素的影响,其方法的有效性受到一定限制;文献的数字化、网络化对引文分析评价方法提出了新的挑战;应该采取一定的改进措施,提高引文分析评价的科学性。     

A content analysis of local television newscasts: Answering the critics

Local television newscasts have been criticized for being too frivolous, but an analysis of San Diego stations indicates more newscast time is devoted to significant issues of the day than to anything else.     

Diversity of body types in network television programming: A content analysis

This study uses content analysis to examine the diversity of men and women's body types on primetime network television programs and afternoon soap operas in order to compare male and female portrayals and to determine if the portrayals of both genders reflect the real population. The results show females are portrayed in a smaller range of body types and are more slender than males; heavier body types are less likely to be in romantic situations, wear revealing clothing, and are older than thin body types; and television body types are generally thinner than the real population.     

Characterizing references from different disciplines: A perspective of citation content analysis

Multidisciplinary cooperation is now common in research since social issues inevitably involve multiple disciplines. In research articles, reference information, especially citation content, is an important representation of communication among different disciplines. Analyzing the distribution characteristics of references from different disciplines in research articles is basic to detecting the sources of referred information and identifying contributions of different disciplines. This work takes articles in PLoS as the data and characterizes the references from different disciplines based on Citation Content Analysis (CCA). First, we download 210,334 full-text articles from PLoS and collect the information of the in-text citations. Then, we identify the discipline of each reference in these academic articles. To characterize the distribution of these references, we analyze three characteristics, namely, the number of citations, the average cited intensity and the average citation length. Finally, we conclude that the distributions of references from different disciplines are significantly different. Although most references come from Natural Science, Humanities and Social Sciences play important roles in the Introduction and Background sections of the articles. Basic disciplines, such as Mathematics, mainly provide research methods in the articles in PLoS. Citations mentioned in the Results and Discussion sections of articles are mainly in-discipline citations, such as citations from Nursing and Medicine in PLoS.     

Censor and sensibility: A content analysis of the television censor's comments

The position of network censor, whatever the title appended to the job, is one of particular sensitivity. To the industry as a whole, he acts to forstall the program material that could cause criticism of the network, station, sponsor or program. A writer who feels that his work has been needlessly mutilated has a different conception of the role of the network censor. The public is largely unaware of his existence.

The pamphlet by the author of the present article, Taste and the Censor in Television (published by the Fund for the Republic, 1959, as an Occasional Paper on the role of the mass media in the free society), was the first major attempt to define the place of the censor. Another article of interest was George Gerbner's “Mental Illness on Television: A Study of Censorship” (Journal of Broadcasting, Vol. III, No. 4, Fall, 1959.)

The article that follows is intended to convey the flavor of the censor's work, and provide an adequate introduction to this important phase of broadcasting self‐regulation and management. It is possibly the only study of the broadcasting executive as a decision‐maker. As such, it should be of extreme interest to anyone in the creative and the business ends of broadcasting.     

Academic library guides for tackling fake news: A content analysis

Built on a pilot study, this study examined how librarians understand fake news and the specific methods or strategies they suggest for detecting fake news by analyzing their guides from academic libraries. A content analysis regarding a total of 21 institutional guides was conducted. The major findings include the following: 1) in the librarians' guides stating their definition of fake news, the two elements of falsity and the intention to mislead were explicitly stated. The other element of bias, however, was presented in only some guides. 2) The sub-elements of clickbait, a decontextualized content and omitted information were inconsistently or barely presented across these guides. 3) Only two institutional libraries put forth the notion of fact in relation to fake news in their guides. 4) All of the guides suggested checklist approaches to detecting fake news or evaluating news sources. Finally, 5) librarians acknowledge the influence of human biases on consuming news. However, psychological factors are minimally presented in most of the guides. This study provides a few suggestions. First, librarians must further clarify the term fake news so that it reflects its multiple layers. Second, librarians must incorporate new strategies, such as lateral reading and click restraints, in combination with a few prioritized elements of a checklist into their guides regarding detecting fake news. Finally, librarians must pay attention to psychological factors more when interpreting facts in their strategies about news sources and fake news.     

图书馆相关法的边际与内容分析——以日本为例

本文阐述了实现图书馆法治,需要有一个图书馆法律保障体系.图书馆相关法是图书馆法律保障体系的重要组成部分,“相关法”是一个非常宽泛的概念,泛指与图书馆相关的一些法规.主要包括图书馆与行政、财政法;图书馆与教育法;图书馆与著作权法;图书馆与地域开发、城市规划法;图书馆与环境保护法;图书馆与残疾人保护法等等.     

网德教育:网络信息安全新课题

人为因素已成为网络安全问题最直接最具影响的重要方面, 种种网德失范行为警示人们必须关注人类的网德建设。重视网德理论研究, 制订具体的网络行为准则和规范, 加强网德教育是深化网德建设的主要措施, 也是网络安全的有力保障。     

A content analysis of job advertisements for digital humanities-related positions in academic libraries

Digital Humanities (DH) is a relatively new and dynamic field that involves a wide range of disciplines. There is a need to identify a set of competencies for professionals in this growing field. The purpose of this study is to assess the current state of required and desirable qualifications and knowledge described in the job advertisements regarding DH-related positions. 72 unique job advertisements between November 2006 and April 2018 were collected from American Library Association JobLIST. Position title; institution types and location; educational background; experience; knowledge and skills; and duties were examined and analyzed. SCI2, VOSviewer and Pajek were used to analyze the word/phrase frequency of the job title and job functions, the clusters of job duties and job requirement. The results of the study offer insights on and have practical applications in DH-related education and training to meet the needs of this field.     

Stylistic variables as related to source credibility: A content analysis approach

The hypothesis was tested that sources rated high in credibility construct messages with more listenability, human interest, vocabulary diversity, realism, and verifiability than sources rated low in credibility. Content analysis techniques applied to messages of high and low credibility sources revealed significant differences in the predicted direction along four of the five factors.     

Examining nonsuicidal self-injury content creation on TikTok through qualitative content analysis

Qualitative content analysis is a methodological approach for the subjective interpretation of data. Using empirical research to illustrate the methodological advantages and disadvantages of qualitative content analysis, this article examines the suitability of qualitative content analysis for the field of LIS and illustrates how the method can be used to inform LIS information practices research of marginalized populations through emerging information and communication technologies (ICTs). Specifically, this article examines the suitability of qualitative content analysis through its application in an ongoing study exploring how individuals who engage in nonsuicidal self-injury circumnavigate algorithmic exclusion in the emerging ICT TikTok. Qualitative content analysis can advance LIS research and practice by refuting deficit thinking and understanding information creation practices in context. Methodological shortcomings relate to the reduction of data and the inability to determine cause-and-effect relationships. The author recommends that researchers supplement qualitative content analysis with additional qualitative approaches to address these limitations.