Αλγόριθμοι ομαδοποίησης χρηστών σε συστήματα σύστασης
User clustering algorithms in recommendation systems
View/
Keywords
Clustering algorithms ; Recommendation systemsAbstract
My thesis examines recommender systems, focusing on various methods and
algorithms used to cluster and recommend data. First, an overview of existing
recommender systems is presented, describing their various categories and functions.
Then, specific clustering algorithms, which are key tools for creating effective
recommender systems, are discussed.
An important part of the paper deals with data preprocessing, where it is discussed
how data is cleaned, transformed and prepared for input to the clustering algorithms.
Finally, we present our own experiments on these algorithms, where we apply the
Pearson correlation, cosine similarity and euclidean distance metrics to the K-means
algorithm. Through this application, we demonstrate the practical use of these metrics
to improve the performance of the algorithm and achieve better clustering results.
Department
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα ΠληροφορικήςNumber of pages
79Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα