Networks often contain groups of nodes that are more densely connected to each other. Community detection is a key method for identifying these structures.
In this update, a new community detection, Leiden Community, has been added.
This algorithm improves on earlier community detection methods and can identify more stable and finer grained community structures.
Text analysis features have also been enhanced.
To make topic analysis easier to interpret, a topic–keyword network map has been added, and document level topic assignments can now be added as attributes to the original dataset.
Additional improvements have been made to visualization features, including edge bundling for network graphs and expanded support for image scaling and export formats.
Apply the latest community detection method to research
such as topic discovery in word networks or political polarization analysis,
and analyze text data more efficiently. | 
