Autor/Autorin: Nkweteyim Denis
Denis Nkweteyim, PhD: Studied Information Science atnUniversity of Pittsburgh; Assistant Lecturer of Computer Sciencenand Manager of IT Centre at the University of Buea, Cameroon;nTeaching/Research: data mining, data bases, programming, datanstructures & algorithms; Interests: design, development,ndeployment, management of information systems.
The information age is characterized by anoverabundance of information but few tools to helpusers of information spaces get just the rightinformation as they navigate these spaces. Currentresearch thus aims to design tools like recommendersystems to personalize the user experience in largeinformation spaces. This book explores factorsinvolved in hyperlink recommender systems design witha view to improving on them or proposing newsolutions. The work makes the following contributions in the area of data mining.Web Page Classification: a new classification metricis developed; Association rule mining: a new apriorialgorithm is developed; Prediction model for userinterests: methodology for extracting from webserver logs, association rules that showcorrelations between user navigation patterns andinteresting web pages, and transformation of therules into collaborative filtering data, widely usedin recommender systems; Clustering: a comparativestudy of the CLARANS, PAM, and CLARA algorithms inhigh dimensional space is presented. The work shouldtherefore be of interest to anyone involved inpersonalization systems and data mining research.
The information age is characterized by an
overabundance of information but few tools to help
users of information spaces get just the right
information as they navigate these spaces. Current
research thus aims to design tools like recommender
systems to personalize the user experience in large
information spaces. This book explores factors
involved in hyperlink recommender systems design with
a view to improving on them or proposing new
solutions. The work makes the following
contributions in the area of data mining.
Web Page Classification: a new classification metric
is developed; Association rule mining: a new apriori
algorithm is developed; Prediction model for user
interests: methodology for extracting from web
server logs, association rules that show
correlations between user navigation patterns and
interesting web pages, and transformation of the
rules into collaborative filtering data, widely used
in recommender systems; Clustering: a comparative
study of the CLARANS, PAM, and CLARA algorithms in
high dimensional space is presented. The work should
therefore be of interest to anyone involved in
personalization systems and data mining research.
Denis Nkweteyim, PhD: Studied Information Science at
University of Pittsburgh; Assistant Lecturer of Computer Science
and Manager of IT Centre at the University of Buea, Cameroon;
Teaching/Research: data mining, data bases, programming, data
structures & algorithms; Interests: design, development,
deployment, management of information systems.
Über den Autor
Denis Nkweteyim, PhD: Studied Information Science atUniversity of Pittsburgh; Assistant Lecturer of Computer Scienceand Manager of IT Centre at the University of Buea, Cameroon;Teaching/Research: data mining, data bases, programming, datastructures & algorithms; Interests: design, development,deployment, management of information systems.
Klappentext
The information age is characterized by anoverabundance of information but few tools to helpusers of information spaces get just the rightinformation as they navigate these spaces. Currentresearch thus aims to design tools like recommendersystems to personalize the user experience in largeinformation spaces. This book explores factorsinvolved in hyperlink recommender systems design witha view to improving on them or proposing newsolutions. The work makes the following contributions in the area of data mining.Web Page Classification: a new classification metricis developed; Association rule mining: a new apriorialgorithm is developed; Prediction model for userinterests: methodology for extracting from webserver logs, association rules that showcorrelations between user navigation patterns andinteresting web pages, and transformation of therules into collaborative filtering data, widely usedin recommender systems; Clustering: a comparativestudy of the CLARANS, PAM, and CLARA algorithms inhigh dimensional space is presented. The work shouldtherefore be of interest to anyone involved inpersonalization systems and data mining research.
