AN APPROACH FOR THE FORMATION OF LEVERAGE COEFFICIENTS-BASED RECOMMENDATIONS IN SOCIAL NETWORK

Lina Tutkute, Rimantas Butleris, Tomas Skersys

Abstract


On the basis of the drawbacks of the current recommender systems additional functionality to the social network, namely, recommender system with management possibility is proposed. In order to implement such functionality to the social network, the following elements of a new method are presented in this paper: a metamodel of the recommendation; an algorithm for leverage coefficients-based recommendations formation; an algorithm of leverage coefficients-based recommendations interpretation. The paper also discusses the possibility to extend social network’s business logic with business rules features and presents some fragments of the prototype that realizes these features.


Full Text: PDF

Print ISSN: 1392-124X 
Online ISSN: 2335-884X