Facebook traffic pattern analytics

C. Chen, J. Iglasias, X. Lin, J. J. Li, P. Morreale, Linda Ness

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

This paper presents a study of applying a new data pattern recognition approach, called Product Coefficients (PCs), to discover patterns of Facebook network and service traffic. It includes three parts: 1) collection of service response time and packages as the initial data set, 2) preprocessing of the data sets and discovers of PCs values as patterns of the data set. 3) displaying of the patterns for simple outlier detection. Additional research is ongoing related to the post processing of PCs values for prediction, e.g. using machine learning to decide if outliers are indicators of a security attack. The key contribution of this work is the application of PCs method to some practical Facebook traffic data.

Original languageEnglish
Title of host publicationProceedings of the 3rd Multidisciplinary International Social Networks Conference, SocialInformatics 2016, Data Science 2016, MISNC, SI, DS 2016
PublisherAssociation for Computing Machinery
ISBN (Print)9781450341295
DOIs
StatePublished - 15 Aug 2016
Event3rd Multidisciplinary International Social Networks Conference, MISNC 2016, 5th ASE International Conference on Social Informatics, SocialInformatics 2016 and 7th ASE International Conference on Data Science, DS 2016 - Union, United States
Duration: 15 Aug 201617 Aug 2016

Publication series

NameACM International Conference Proceeding Series

Conference

Conference3rd Multidisciplinary International Social Networks Conference, MISNC 2016, 5th ASE International Conference on Social Informatics, SocialInformatics 2016 and 7th ASE International Conference on Data Science, DS 2016
Country/TerritoryUnited States
CityUnion
Period15/08/1617/08/16

Keywords

  • Machine learning
  • Outliers
  • Predictive analytics
  • Statistical patterns

Fingerprint

Dive into the research topics of 'Facebook traffic pattern analytics'. Together they form a unique fingerprint.

Cite this