Unique unto itself: The records of the Federal Bureau of Investigation 1908 to 1945

This paper addresses several misunderstandings about the nature of the FBI's Central Records System. Although depicted in popular culture and much of FBI historiography as a tool for the Cold War suppression of American radicals, the FBI's records system grew out of Department of Justice systems already in place before J. Edgar Hoover became Director in 1924. The system grew in size and complexity according to the demands placed on it by the changing priorities of the FBI's criminal and intelligence missions, guided by a “progressive” drive for efficiency through centralization.     

Character N-Gram Tokenization for European Language Text Retrieval

The Cross-Language Evaluation Forum has encouraged research in text retrieval methods for numerous European languages and has developed durable test suites that allow language-specific techniques to be investigated and compared. The labor associated with crafting a retrieval system that takes advantage of sophisticated linguistic methods is daunting. We examine whether language-neutral methods can achieve accuracy comparable to language-specific methods with less concomitant software complexity. Using the CLEF 2002 test set we demonstrate empirically how overlapping character n-gram tokenization can provide retrieval accuracy that rivals the best current language-specific approaches for European languages. We show that n = 4 is a good choice for those languages, and document the increased storage and time requirements of the technique. We report on the benefits of and challenges posed by n-grams, and explain peculiarities attendant to bilingual retrieval. Our findings demonstrate clearly that accuracy using n-gram indexing rivals or exceeds accuracy using unnormalized words, for both monolingual and bilingual retrieval.     

Scholarly communication in the digital environment: what do authors want?

This article reports on a large‐scale international survey of authors' perception and experience of the journals system conducted by ciber in association with National Opinion Polls (NOP). It explores the factors that inform authors' decisions where to publish and, in particular, which groups of readers they perceive to be most important. It probes readership behaviour and the values that underlie authors' attitudes towards copyright and emerging business models, notably open access. It is concluded that many aspects of author behaviour are highly conservative and that a significant shift towards open access is, in the short to medium term, highly unlikely.     

迈向数字化企业

当您每年要向中国数千个地方发送1600万个包裹时,很多问题都可能出现。只要有百分之一的包裹被标示错误,就会有16万个包裹迷失目的地,势必引来大群客户的抱怨。     

Macroappraisal Kiwi Style: Reflections on the Impact and Future of Macroappraisal in New Zealand

Both macroappraisal and the Australian records continuum-based DIRKS methodology have been influential in the New Zealand debate on appraisal in recent years. The primary influence of macroappraisal has been in the area of prioritisation of appraisal work. This paper considers New Zealand thinking on prioritisation, and the influences of risk management and functional analysis on this issue. A lack of agreement on the purpose of appraisal in the professional literature is noted, and some personal suggestions are offered on a model taking elements from macroappraisal and other methodologies.     

Analysis of Statistical Question Classification for Fact-Based Questions

Question classification systems play an important role in question answering systems and can be used in a wide range of other domains. The goal of question classification is to accurately assign labels to questions based on expected answer type. Most approaches in the past have relied on matching questions against hand-crafted rules. However, rules require laborious effort to create and often suffer from being too specific. Statistical question classification methods overcome these issues by employing machine learning techniques. We empirically show that a statistical approach is robust and achieves good performance on three diverse data sets with little or no hand tuning. Furthermore, we examine the role different syntactic and semantic features have on performance. We find that semantic features tend to increase performance more than purely syntactic features. Finally, we analyze common causes of misclassification error and provide insight into ways they may be overcome.     

Google Scholar's Coverage of the Engineering Literature 10 years Later

In 2008 Meier and Conkling first tested Google Scholar's coverage of the engineering literature against citations gathered from the Compendex database. Since that time, other studies have used the same methodology and found improvement in Google Scholar's coverage. This study uses engineering dissertations from Proquest Dissertations & Theses to create a data set of citations for the comparison of fee-based databases, Compendex and Scopus, against Google Scholar. From 1950 to 2017 Google Scholar outperformed both Compendex and Scopus in discoverability of citations in nine engineering subjects. These results have implications for collection management and information literacy program planning for librarians.