A Model for the Recommender System Creation and Collaborative System Filtering Using Data Management |
Katarina Jovanovic and Milena Stankovic Faculty of Electronic Engineering at University of Niš Aleksandra Medvedeva 14, Niš 1800 Serbia |
Abstract: In recommender system, data management techniques are important. Using collaborative filtering and zero suppressed binary decision diagram, we have developed a model for the recommender system creation. We have used the diagrams of the binary decision diagrams to analyse the issues of the large data sets. To prepare the user product matrix for preserving the user behaviour of the collaborative filtering system, binary decision diagrams are required. We in this work have developed recommendation system for a class of users which provide the requirements for each user based on the diagram traversal. We found that when large number of users and products are available, this method of the recommender system will be useful. |
Keywords: Collaborative Filtering, Zero-suppressed BDD, Tuple Histogram, Pattern Histogram A Model for the Recommender System Creation and Collaborative System Filtering Using Data Management |
DOI:https://doi.org/10.6025/jio/2021/11/3/72-77 |
Full_Text   PDF 547 KB   Download:   220  times |
References: [1] Asanov, D. Algorithms and Methods in Recommender System, Berlin Institute of Technology, Berlin, Germany. |