Andreas W. Neumann

Others Also Use: A Robust Recommender System for Scientific Libraries

Scientific digital library systems are a very promising application area for value-added expert advice services. Such systems could significantly reduce the search and evaluation costs of information products for students and scientists. This holds for pure digital libraries as well as for traditional scientific libraries with online public access catalogs (OPAC). In this contribution we first outline different types of recommendation services for scientific libraries and their general integration strategies. Then we focus on a recommender system based on log file analysis that is fully operational within the legacy library system of the Universitat Karlsruhe (TH) since June 2002. Its underlying mathematical model, the implementation within the OPAC, as well as the first user evaluation is presented.

An integration strategy for distributed recommender services in legacy library systems

Scientific library systems are a very promising application area for recommender services. Scientific libraries could easily develop customer-oriented service portals in the style of amazon.com. Students, university teachers and researchers can reduce their transaction cost (i.e. search and evaluation cost of information products). For librarians, the advantage is an improvement of the customer support by recommendations and the additional support in marketing research, product evaluation, and book selection. In this contribution we present a strategy for integrating a behavior-based distributed recommender service in legacy library systems with minimal changes in the legacy systems.

Recommender Services in Scientific Digital Libraries

In this article we give a survey of the current practice and state-of-the-art of recommender services in scientific digital libraries. With the notable exception of amazon.com and CiteSeer which do not qualify as proper scientific libraries our survey revealed that in scientific libraries recommender services are still not in wide use — despite the considerable benefits they offer for students and scientists. This fact can at least partially be explained by mechanism design problems which exist for the basic types of recommender systems and decreased funding for scientific libraries. Next, we present the principles of four recommender services developed at the Universitat Karlsruhe (TH), namely the explicit review and rating service of the library of the Universitat Karlsruhe (TH), the implicit basic “Others also searched …” service (BibTip) of the library of the Universitat Karlsruhe (TH), the prototypes of its small sample and its adaptive variant. A discussion of the current industry trend towards social spaces and societies and its potential for scientific digital libraries concludes this contribution.

Motivating and supporting user interaction with recommender systems

This contribution reports on the introduction of explicit recommender systems at the University Library of Karlsruhe. In March 2006, a rating service and a review service were added to the already existing behavior-based recommender system. Logged-in users can write reviews and rate all library documents (books, journals, multimedia, etc.); reading reviews and inspecting ratings are open to the general public. A role system is implemented that supports the submission of different reviews for the same document from one user to different user groups (students, scientists, etc.). Mechanism design problems like bias and free riding are discussed, to address these problems the introduction of incentive systems is described. Usage statistics are given and the question, which recommender system supports which user needs best, is covered. Summing up, recommender systems are a way to combine the support of library user interaction with information access beyond catalog searches.

Building Recommendations from Random Walks on Library OPAC Usage Data

In this contribution we describe a new way of building a recommender service based on OPAC web-usage histories. The service is based on a clustering approach with restricted random walks. This algorithm has some properties of single linkage clustering and suffers from the same deficiency, namely bridging. By introducing the idea of a walk context (see Franke and Thede (2005) and Franke and Geyer-Schulz (2004)) the bridging effect can be considerably reduced and small clusters suitable as recommendations are produced. The resulting clustering algorithm scales well for the large data sets in library networks. It complements behavior-based recommender services by supporting the exploration of the revealed semantic net of a library network’s documents and it offers the user the choice of the trade-off between precision and recall. The architecture of the behavior-based system is described in Geyer-Schulz et al. (2003).

Applying Small Sample Test Statistics for Behavior-based Recommendations

This contribution reports on the development of small sample test statistics for identifiying recommendations in market baskets. The main application is to lessen the cold start problem of behavior-based recommender systems by faster generating quality recommendations out of the first small samples of user behavior. The derived methods are applied in the area of library networks but are generally applicable in any consumer store setting. Analysis of market basket size at different organisational levels of German research library networks reveals that at the highest network level market basket size is considerably smaller than at the university level. The overall data volume is considerably higher. These facts motivate the development of small sample tests for the identification of non-random sample patterns. As in repeat-purchase theory the independent stochastic processes are modelled. The small sample tests are based on modelling the choice-acts of a decision maker completely without preferences by a multinomial model and combinatorial enumeration over a series of increasing event spaces. A closed form of the counting process is derived.

RecoDiver: Browsing behavior-based recommendations on dynamic graphs

Various kinds of recommendation services open to the general public have recently been integrated into the website of the University Library of Karlsruhe as a test bed for information providers and e-commerce alike. This contribution reports on the development of RecoDiver, a graph-based user interface for behavior-based recommender systems. A Java applet integrated into the librarys online catalog dynamically displays recommended further documents in a clickable graph centered around the document of interest to the user. A local view of the complete graph of recommendations is presented in a radial tree layout based on a minimum spanning tree with animated graph transitions featuring interpolations by polar coordinates to avoid crisscrossings. Further graph search tools like a selectable histogram of years of publication are available as well. This article portrays the user interface as well as the distributed web service architecture behind it and features an evaluation by user surveys showing the preference of users compared to the common lists of recommended items.

GIRAN: A Dynamic Graph Interface to Neighborhoods of Related Articles

This contribution reports on the development of GIRAN (Graph Interface to Related Article Neighborhoods), a distributed web application featuring a Java applet user front-end for browsing recommended neighborhoods within the network of Wikipedia articles. The calculation of the neighborhood is based on a graph analysis considering articles as nodes and links as edges. The more the link structure of articles is similar to the article of current interest, the more they are considered related and hence recommended to the user. The similarity strength is depicted in the graph view by means of the width of the edges. A Java applet dynamically displays the neighborhood of related articles in a clickable graph centered around the document of interest to the user. The local view moves along the complete article network when the user shows a new preference by clicking on one of the presented nodes. The path of selected articles is stored, can be displayed within the graph, and is accessible by the user; the content of the article of current interest is displayed next to the graph view. The graph of recommended articles is presented in a radial tree layout based on a minimum spanning tree with animated graph transitions featuring interpolations by polar coordinates to avoid crisscrossings. Further graph search tools and filtering techniques like a selectable histogram of Wikipedia categories and a text search are available as well. This contribution portrays the graph analysis methods for thinning out the graph, the dynamic user interface, as well as the service-oriented architecture of the application back-end.