TY  - JOUR
T1  - Friend Recommendation System based on Modeling the
Communities using Naive Bayes
AU - Nawaf, Huda N. AU - Al-Hameed, Wafaa. AU - Rasheed Jouda, Najah 
JO  - Journal of Engineering and Applied Sciences
VL  - 12
IS  - 16
SP  - 4156
EP  - 4160
PY  - 2017
DA  - 2001/08/19
SN  - 1816-949x
DO  - jeasci.2017.4156.4160
UR  - https://makhillpublications.co/view-article.php?doi=jeasci.2017.4156.4160
KW  - Friend recommendation system
KW  -Naive Bayes
KW  -centrality measures
KW  -Ego networks
KW  -communities
KW  -Twitter
AB  - Popular social networks sites such as Facebook and Twitter are still growing significantly. In this
regard, a recommender system can be used to provide user experiences. In this study, we try modeling the
online communities using naive bayes model. More specifically, the core of this work is modeling the user’s
past friends by taking into account the centrality measures and the latest friends. The two real datasets
Facebook-Ego and Twitter are consider as a test bed for our proposed system then precision and recall
measures have been applied to evaluate the accuracy of the system. In addition, a new metric, namely the R<sub>top-list</sub>
metric is suggested to express the accuracy of prediction. In sum, the empirical results foster the efficiency of
the proposed system.
ER  -