James A. Stemper

Automatically building research reading lists

All new researchers face the daunting task of familiarizing themselves with the existing body of research literature in their respective fields. Recommender algorithms could aid in preparing these lists, but most current algorithms do not understand how to rate the importance of a paper within the literature, which might limit their effectiveness in this domain. We explore several methods for augmenting existing collaborative and content-based filtering algorithms with measures of the influence of a paper within the web of citations. We measure influence using well-known algorithms, such as HITS and PageRank, for measuring a nodes importance in a graph. Among these augmentation methods is a novel method for using importance scores to influence collaborative filtering. We present a task-centered evaluation, including both an offline analysis and a user study, of the performance of the algorithms. Results from these studies indicate that collaborative filtering outperforms content-based approaches for generating introductory reading lists.

Developing a Model To Provide Digital Reference Services.

Developing digital reference services within a large library system requires managing fundamental organizational issues before applying technological strategies. Using the development of library services for distance learners as a catalyst, the University of Minnesota‐Twin Cities Libraries have implemented an organizational model for providing digital reference services to all students, faculty and staff who access the library remotely. The resulting service, InfoPoint, has the dual role of providing reference service and making referrals to over 30 information service units in the library system. Details the planning and implementation process by which the libraries developed this centrally coordinated digital reference service within a historically decentralized environment. Initial conclusions based on the service’s first two years of operation are presented. In conclusion discusses organizational change issues and the value of digital reference services in the digital library.

Usage Statistics for Electronic Journals: An Analysis of Local and Vendor Counts

Abstract Vendor-supplied e-resource statistics are often unavailable, unreliable, or not comparable across vendors. This study compared locally generated usage statistics to those supplied by four major publishers, and analyzed the resulting patterns of use. The additional information provided by vendor statistics was assessed to see how it might be utilized selectively to provide a better understanding of the importance of individual titles. The local statistics for all titles provided by the four publishers were then compared. A strong similarity between the two datasets was found, supporting the position that local statistics are a viable alternative to vendor statistics. Another finding was that the 80/20 rule is closer to 80/30 in the online environment. Some of the issues for subject librarians and electronic resources librarians are discussed.

Techlens: a researcher's desktop

Rapid and continuous growth of digital libraries, coupled with brisk advancements in technology, has driven users to seek tools and services that are not only customized to their specific needs, but are also helpful in keeping them stay abreast with the latest developments in their field. TechLens is a recommender system that learns about its users through implicit feedback, builds correlations among them, and uses that information to generate recommendations that match the users profile. It gives users control over which parts of their profile of known citations are used in forming recommendations for new articles. This demonstration is a prototype that showcases some of the tools and services that TechLens offers to the users of digital libraries.

E-Journal Metrics for Collection Management: Exploring Disciplinary Usage Differences in Scopus and Web of Science

Objective – The purpose was to determine whether a relationship exists between journal downloads and either faculty authoring venue or citations to these faculty, or whether a relationship exists between journal rankings and local authoring venues or citations. A related purpose was to determine if any such relationship varied between or within disciplines. A final purpose was to determine if specific tools for ranking journals or indexing authorship and citation were demonstrably better than alternatives. Methods – Multiple years of journal usage, ranking, and citation data for twelve disciplines were combined in Excel, and the strength of relationships were determined using rank correlation coefficients. Results – The results illustrated marked disciplinary variation as to the degree that faculty decisions to download a journal article can be used as a proxy to predict which journals they will publish in or which journals will cite faculty’s work. While journal access requests show moderate to strong relationships with the journals in which faculty publish, as well as journals whose articles cite local faculty, the data suggest that Scopus may be the better resource to find such information for these journals in the health sciences and Web of Science may be the better resource for all other disciplines analyzed. The same can be said for the ability of external ranking mechanisms to predict faculty publishing behaviours. Eigenfactor is more predictive for both authoring and citing-by-others across most of the representative disciplines in the social sciences as well as the physical and natural sciences. With the health sciences, no clear pattern emerges. Conclusion – Collecting and correlating authorship and citation data allows patterns of use to emerge, resulting in a more accurate picture of use activity than the commonly used cost-per-use method. To find the best information on authoring activity by local faculty for subscribed journals, use Scopus. To find the best information on citing activity by faculty peers for subscribed titles use Thomson Reuters’ customized Local Journal Use Reports (LJUR), or limit a Web of Science search to local institution. The Eigenfactor and SNIP journal quality metrics results can better inform selection decisions, and are publicly available. Given the trend toward more centralized collection development, it is still critical to obtain liaison input no matter what datasets are used for decision making. This evidence of value can be used to defend any local library “tax” that academic departments pay as well as promote services to help faculty demonstrate their research impact.