|
|
 |
| |
|
|
|
Binomial Regressive Influence Behavior Ranking for Virtual Community Formation in Social Network |
|
|
|
PP: 935-943 |
|
|
|
doi:10.18576/amis/130606
|
|
|
|
Author(s) |
|
|
|
R. Gnanakumari,
P. Vijayalakshmi,
|
|
|
|
Abstract |
|
|
| A Social Network (SN) is a website which permits people to share the
data about their personal or business endeavor to form the virtual
community. Due to communication between the users in the SN, a similar
users behavior identification causes a fundamental issue. The existing
techniques still encounter the problem of identifying similar behavior
accurately. Therefore, Statistic Dice Similarity Based Probabilistic
Binomial Regression and Ranking (SDS-PBRR) method introduced. First, the
similarity value between the users behaviors is calculated using Statistic
Dice Similarity Coefficient (SDSC). Second, Probabilistic Binomial
Regression Analysis is carried out to evaluate the similarity value and to
classify the users as Influencing Behavior (IB) or Non-Influencing Behavior
(N-IB) minimum error rate in the SN. Last, firefly algorithm is applied to
perform a ranking process for discovering the level of IB users in the SN.
The simulation results show that SDS-PBRR method increases the True Positive
Rate (TPR) and minimizes the False Positive Rate (FPR) as well as execution
time. |
|
|
|
|
|
|
|
