Gefi Clustering (Louven Method)

Question

I started working with gephi to help me display a dataset. The data set contains:

tags (terms for a specific image) as nodes

Google normalized gathering Google Distance between these tags as ribs with weight (between 0 and 1)

Each tag connects to all other tags if they both belong to the same image. Therefore, I have one cluster of nodes and edges for each image.

Now I have imported this dataset into gephi in the following format:

nodes: id, label

ribs: target, source, weight (0 to 1)

Like 500 knots and 6,000 ribs.

Now my problem is that after importing all these nodes and edges, the graph looks grouped without real order. Each cluster of each image is mixed with other clusters of other images. Now, using the modularization algorithm as the Partitition algorithm (which should use the Louvain method), the graph becomes color, each color is an image. Now I can break this mess using the Force Atlas 2 layout.

Now I have a color chart with something like 15 clusters (each cluster represents 1 image)

Now I want to group these clusters again using tags (nodes) according to their normalized google distance (weight of edges), which should be tags that are somewhat equal in value.

I hope you guys understand what I want to achieve. I can also download picutre to clarify this.

thanks a lot

Answer 1

I was able to solve my problem. I had to import each of these 15 clusters on my own. Thus, I could use the modularity method only for those few.

Answer 2

I do not think you can do this with the standard version of Gephi. You will need to develop a plugin to implement the very last step of your process.

Gephi is good for visualizing and viewing graphs, but (at the moment) there are more complete tools when it comes to processing topological properties. for example, the igraph library (available in C, R, and python) might be more suitable for you. And note that you can use a file format compatible with both Gephi and igraph, which allows you to use both tools for the same data.