Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Link prediction in multiplex online social networks.

Mahdi Jalili1, Yasin Orouskhani2, Milad Asgari3

  • 1School of Engineering , RMIT University , Melbourne, Victoria , Australia.

Royal Society Open Science
|April 8, 2017
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Diffusion-induced instabilities promote cooperation in eco-evolutionary networks.

Proceedings of the National Academy of Sciences of the United States of America·2026
Same author

Symmetry-protected delay spectroscopy in oscillator networks.

Chaos (Woodbury, N.Y.)·2026
Same author

Predictive factors for poor mobilization in autologous stem cell transplant: a multivariate model.

Hematology, transfusion and cell therapy·2026
Same author

Quercetin-loaded cellulose nanofibers improve memory, learning, and attenuate endoplasmic reticulum stress in a rat model of Alzheimer's disease.

Scientific reports·2026
Same author

Rumor propagation and supervision during confrontation: An importance-driven SIRQS network model.

Chaos (Woodbury, N.Y.)·2026
Same author

Unintended consequences of well-intended interventions.

Proceedings of the National Academy of Sciences of the United States of America·2026

This study introduces a meta-path-based algorithm for link prediction in multiplex social networks, like Twitter and Foursquare. Integrating cross-layer information significantly boosts prediction accuracy, achieving 89% with SVM.

Area of Science:

  • Social Network Analysis
  • Data Mining
  • Machine Learning

Background:

  • Online social networks are integral to modern society, influencing relationship dynamics.
  • Link prediction in social networks has diverse applications, including recommendations and connection discovery.
  • Real-world social networks often exist in multiple layers, such as across different platforms.

Purpose of the Study:

  • To investigate link prediction in multiplex social networks.
  • To develop and evaluate a meta-path-based algorithm for predicting links using cross-layer information.
  • To assess the effectiveness of different classifiers in this task.

Main Methods:

  • Constructed a multiplex network using Twitter and Foursquare data from the same users.
Keywords:
complex networkslink predictionmachine learningsigned networkssocial networks

Related Experiment Videos

  • Developed a meta-path-based algorithm leveraging connectivity information from both layers.
  • Employed Naive Bayes, Support Vector Machines (SVM), and K-nearest neighbour classifiers for link prediction.

    • Main Results:

    • Cross-layer information significantly improved link prediction performance, even with low correlation between network layers.
    • The Support Vector Machines (SVM) classifier achieved the highest prediction accuracy.
    • An average accuracy of 89% was obtained using the SVM classifier.

    Conclusions:

    • Multiplex network analysis enhances link prediction accuracy compared to single-layer approaches.
    • Integrating data from different social platforms provides valuable insights for understanding user connections.
    • The proposed meta-path-based method with SVM is effective for link prediction in complex social network structures.