Innovative and efficient cluster analysis of networks with textual edges

Linkage allows you to cluster the nodes of networks with textual edges while identifying topics which are used in communications. You can analyze with Linkage networks such as email networks or co-authorship networks. Linkage allows you to upload your own network data or to make requests on scientific databases (Arxiv, Pubmed, HAL).




  How does Linkage work ?

Linkage is built upon a sound statistical model for networks with textual edges and implement an innovative and efficient inference algorithm to fit the model on your data. Model selection allows to find in a fully automatic way the best number of groups and topics.

  Upload and manage your data securely

Upload all or part of your data on the platform to analyze them with Linkage. You will keep full control on the data you upload and only you will be able to access them.

  Focus on data and interpretation

Minimum configuration is required to use Linkage since it selects the most sensible parameters for the data you provide. No scientific background is required to start working and get results. Advanced configuration options are also available if you need specific setups.

  Visualize and export the results

Linkage also provides advanced visualization tools, based on the specific features of the statistical modeling. Linkage finally allows to export as CSV files the clustering results obtained on your data for further processing.

  The statistical method behind the platform

The methodology implemented is partly related to an article published in the journal « Statistics and Computing ». The reference to cite in case of academic use of the platform is « C. Bouveyron, P. Latouche and R. Zreik, The Stochastic Topic Block Model for the Clustering of Networks with Textual Edges, Statistics and Computing, in press, 2017. DOI: 10.1007/s11222-016-9713-7 ».

  The people behind

Linkage is developed by an academic team:


                                             

Terms and conditions         Credits