|
Activity Number:
|
530
- Contributed Poster Presentations: Section on Statistics in Marketing
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Marketing
|
|
Abstract #307104
|
|
|
Title:
|
Classification of Social Media Users Through Generalized Multilevel Functional Model
|
|
Author(s):
|
Anthony Weishampel* and Bill Rand and Ana-Maria Staicu
|
|
Companies:
|
North Carolina State University and North Carolina State University and North Carolina State University
|
|
Keywords:
|
Generalized Functional Data;
Social Media;
Multilevel Functional Data;
Principal Component Estimation
|
|
Abstract:
|
Social Media provides a far-reaching platform for in?uencers to spread their content. Firm social media and word of mouth campaigns are susceptible to malicious social media users. Being able to detect various types of users is vital to mitigate their effects. This research builds a classification method based on the user’s behavior and other covariate information. A user’s behavior is treated as a binary time series, indicating times of activity. We consider a generalized multilevel functional model for the response profile. This model separates the user-specific variation from the day within user variation and from the mean trend, while accounting for additional covariate effects. The user-specific and day within user trends are estimated through functional principal components analyses. Classification of the users is accomplished through analyzing the user-specific trends. The results are compared to the existing social media classification approaches and other time dependent models.
|
Authors who are presenting talks have a * after their name.
