Hierarchical clustering of a Finnish newspaper article collection with graded relevance assessments

Search facilitated with agglomerative hierarchical clustering methods was studied in a collection of Finnish newspaper articles (N = 53,893). To allow quick experiments, clustering was applied to a sample (N = 5,000) that was reduced with principal components analysis. The dendrograms were heuristically cut to find an optimal partition, whose clusters were compared with each of the 30 queries to retrieve the best-matching cluster. The four-level relevance assessment was collapsed into a binary one by (A) considering all the relevant and (B) only the highly relevant documents relevant, respectively. Single linkage (SL) was the worst method. It created many tiny clusters, and, consequently, searches enabled with it had high precision and low recall. The complete linkage (CL), average linkage (AL), and Ward's methods (WM) returned reasonably-sized clusters typically of 18–32 documents. Their recall (A: 27–52%, B: 50–82%) and precision (A: 83–90%, B: 18–21%) was higher than and comparable to those of the SL clusters, respectively. The AL and WM clustering had 1–8% better effectiveness than nearest neighbor searching (NN), and SL and CL were 1–9% less efficient that NN. However, the differences were statistically insignificant. When evaluated with the liberal assessment A, the results suggest that the AL and WM clustering offer better retrieval ability than NN. Assessment B renders the AL and WM clustering better than NN, when recall is considered more important than precision. The results imply that collections in the highly inflectional and agglutinative languages, such as Finnish, may be clustered as the collections in English, provided that documents are appropriately preprocessed.     

Can constrained relevance feedback and display strategies help users retrieve items on mobile devices?

Searching online information resources using mobile devices is affected by small screens which can display only a fraction of ranked search results. In this paper we investigate whether the search effort can be reduced by means of a simple user feedback: for a screenful of search results the user is encouraged to indicate a single most relevant document. In our approach we exploit the fact that, for small display sizes and limited user actions, we can construct a user decision tree representing all possible outcomes of the user interaction with the system. Examining the trees we can compute an upper limit on relevance feedback performance. In this study we consider three standard feedback algorithms: Rocchio, Robertson/Sparck-Jones (RSJ) and a Bayesian algorithm. We evaluate them in conjunction with two strategies for presenting search results: a document ranking that attempts to maximize information gain from the user’s choices and the top-D ranked documents. Experimental results indicate that for RSJ feedback which involves an explicit feature selection policy, the greedy top-D display is more appropriate. For the other two algorithms, the exploratory display that maximizes information gain produces better results. We conducted a user study to compare the performance of the relevance feedback methods with real users and compare the results with the findings from the tree analysis. This comparison between the simulations and real user behaviour indicates that the Bayesian algorithm, coupled with the sampled display, is the most effective. Extended version of “Evaluating Relevance Feedback Algorithms for Searching on Small Displays, ” Vishwa Vinay, Ingemar J. Cox, Natasa Milic-Frayling, Ken Wood published in the proceedings of ECIR 2005, David E. Losada, Juan M. Fernández-Luna (Eds.), Springer 2005, ISBN 3-540-25295-9     

Employee mistreatment and muted voices in the culturally diverse workplace

This study examines mistreatment through the perspectives of employees with different cultural backgrounds and positions in order to understand the ways in which some voices are muted and others are privileged. Mistreatment is interactional, distributive, procedural, or systemic abuse of employees taking place at both interpersonal and institutional levels on the basis of cultural diversity. Narrative analysis of semi-structured interviews with employees of a large research and development organization revealed three types of muted narratives used to respond to mistreatment (muted-but-engaged, angrily disengaged, and resigned) as well as one type of privileged response. The narratives provided evidence of two processes by which voices became muted: repeated silencing over time and silencing through ambiguity of policies.     

Indianapolis I3: the third generation Integrated Advanced Information Management Systems.

In 2001, the Regenstrief Institute for Health Care and the Indiana University School of Medicine (IUSM) began an IAIMS planning effort to create a vision and a tactical plan for the first Integrated Advanced Information Management Systems (IAIMS) implementation to cross a large area and include unaffiliated institutions. A number of elements made this planning effort unique. Among these elements were the existence of a network infrastructure that supported the Indianapolis Network for Patient Care, the existence of a mature medical informatics program at the Regenstrief Institute, and the existence of a wide-area knowledge network fostered by the IUSM libraries. However, the leadership for a strong information technology role in the IUSM that could promote collaboration in support of education and research across the diverse Indianapolis hospital systems had been lacking. By bringing together various groups, each with a commitment to improve health care quality and public health across the Indianapolis metropolitan area, regardless of individual institutional affiliation, the strategic directions for I3-Indianapolis IAIMS Initiative have been defined and the foundations for a third generation IAIMS construct have been laid in Indianapolis, Indiana.     

Web usage mining at an academic health sciences library: an exploratory study.

OBJECTIVES: This paper explores the potential of multinomial logistic regression analysis to perform Web usage mining for an academic health sciences library Website. METHODS: Usage of database-driven resource gateway pages was logged for a six-month period, including information about users' network addresses, referring uniform resource locators (URLs), and types of resource accessed. RESULTS: It was found that referring URL did vary significantly by two factors: whether a user was on-campus and what type of resource was accessed. CONCLUSIONS: Although the data available for analysis are limited by the nature of the Web and concerns for privacy, this method demonstrates the potential for gaining insight into Web usage that supplements Web log analysis. It can be used to improve the design of static and dynamic Websites today and could be used in the design of more advanced Web systems in the future.     

Forschungs- und Entwicklungsprojekte Virtuelle Maschinen: zSeries- und S/390-Partitionierung

Der gleichzeitige Betrieb mehrerer Gast-Betriebssysteme auf einem einzigen physischen Rechner unter einem Host-Betriebssystem ist eine leistungsfähige moderne Entwicklung. Bekannte Beispiele sind VMware für die IA32-Architektur sowie das Betriebssystem z/VM und die PR/SM-LPAR Einrichtungen der zSeries-Architektur. Die Nutzung eines Betriebssystems als Gast bedingt einen Leistungsverlust. Die als Partitionierung bezeichnete Zuordnung von Systemressourcen zu den einzelnen Gast-Betriebssysteme ist schwierig, wenn eine dynamische Anpassung an sich ändernde Lastprofile erforderlich ist. Diese Probleme lassen sich mittels Erweiterungen der Hardwarearchitektur adressieren, sowie durch Softwarestrukturen, welche diese Erweiterungen nutzen. Die Erweiterungen der Hardwarearchitektur gehen über das hinaus, was auf heutigen Rechnerarchitekturen wie IA32 oder Mips verfügbar ist. Der vorliegende Beitrag erläutert den optimalen Betrieb von Gast-Betriebssystemen und die begleitenden Partitionierungsmöglichkeiten auf der zSeries-Plattform und beschreibt die zusätzlichen Hardware- und Software-Einrichtungen, welche dies ermöglichen.     

Grundlagen des Burrows-Wheeler-Kompressionsalgorithmus

Der Burrows-Wheeler-Kompressionsalgorithmus hat als ein universelles Kompressionsverfahren in den letzten Jahren aufgrund seiner hervorragenden Kompressionsraten und hohen Verarbeitungsgeschwindigkeiten eine bemerkenswerte Beachtung erfahren. Das Herzstück des Algorithmus stellt die sogenannte Burrows-Wheeler-Transformation dar. Es handelt sich bei dieser Transformation um eine Permutation der Eingabedaten, welche Zeichen mit ähnlichem Kontext nahe beieinander anordnet. Die Transformation sowie die zugehörige Rücktransformation werden zusammen mit den weiteren Stufen des Algorithmus in einer gegenüber der Standardversion verbesserten Variante vorgestellt und die Ergebnisse mit denen anderer Datenkompressionsalgorithmen verglichen.     

Scholarly communication trends — voices from the vortex: a summary of specialist opinion

The views of specialists (academics, librarians, publishers and information scientists) on trends in scholarly communications were canvassed by means of structured questions. A greater commonality of views was found than might have been expected.