Projecten en Activiteiten - UHasselt

Course in Documentation and Library Science - UA

Prof. Dr. L. Egghe (Chief Librarian of Universiteit Hasselt) teaches two courses in this program. The course descriptions are given below. Each course takes 30 hours and is taught once every two years. For more information : contact Prof. Dr. L. Egghe. His email is : leo.egghe@uhasselt.be

Quantitative Methods in the Information Area
Prof. Dr. L. Egghe

The general goal of the course is to inform students on the different statistical and mathematical methods that can be applied to libraries and other information centres. The course contains information on the methodology in statistics, citation analysis, queueing theory and informetric laws.
The main goal of the statistical part is to teach the students how to collect and interpret data on the performance of libraries and other information centres. In this way management decisions can be taken in a professional way. Time is devoted to graphical aspects (incl. time evolutions) as well as to the preparation of reports (such as annual reports). Special attention is given to the execution and interpretation of samples and how they can be used to predict certain parameters for the entire library or information centre. Many exercises are given dealing with "real life" problems.
We also deal with multivariable statistics in which it is the purpose to make "atlasses" of related objects such as libraries, journals or authors. They have applications in the science of science and in science policy, such as citation analysis. The latter is a separate chapter that also deals with the JCR (Journal Citation Reports - a statistical product of Thomson-ISI, the Institute for Scientific Information) and with derived parameters such as growth and obsolescence (ageing) of the literature.
Queueing theory predicts how long items (such as library users) will have to wait in line before they receive the requested service. We e.g. study circulation desks with one or several servers and examine the effect of splitting up service centres.
Informetric laws (such as the ones of Lotka, Bradford, Zipf, Mandelbrot, Leimkuhler) are the basis for explaining certain regularities in information science, hereby using methods of the exact sciences. Derived properties are: success-breeds-success, the so-called 80/20 rule and aspects of inequality (concentration theory).
In the course one uses a scientific pocket calculator as well as a statistical software package. The students must be able to understand the theoretical aspects as well as to solve exercises. For the latter, the course itself can be used.

Documentary Systems: Retrieval
Prof. Dr. L. Egghe

The main goal of this course is to teach the students the logical aspects of documentary systems, how they are constituted as well as how they are used in order to retrieve information.
Different documentary system types are discussed. Then attention is given to record and file structures as well as to a classification of information retrieval models.
Search strategies consume several chapters. We start with aspects of query formulation. Based on the inverted file structure one discusses Boolean searches, proximity searches (various types), limitations to parts of the record and searches based on thesauri. Methodological aspects of retrieval are discussed. All these techniques are illustrated by live demonstrations in several documentary systems (e.g. DIALOG, ERL, WWW, ...). We also show how search results can be seen as well as saved.
Next we devote our attention to evaluation aspects of documentary systems (e.g. recall and precision) and how indexing and retrieval techniques have influence on this. Special attention is given to the difficult problem of evaluation of browsers on the Internet.
A second part of the course deals with information theory and coding theory. First of all a rigorous definition of "bit" is given, hereby using the entropy machinery (i.e. aspects of Shannon-Weaver theory). Applications in information retrieval and on automatic indexing are given.
Next coding theory deals with error detection and correction as well as compression of texts (Huffman compression again using entropy). In this connection the theory of ISBN, ISSN and CODEN is given and exercises made. The course closes with some aspects of public key cryptography.
The written exam emphasises abilities of logical thinking as is needed in search formulations. Practical problems are posed.