I produced a semantic network after the following operations:
- terms extraction
- light clean-up of the list
- indexing of the resulting “Terms” list
- network mapping “Terms” x “Terms” (500 nodes, distributional)
Why a term’s “weight” is not the same depending on whether I consult the “weight” column of the clusters.csv table provided in maps_output.zip\maps\maps_export_format\clusters.csv, or whether I refer (by clicking on the relevant node) to the “weigth” data in the “Explore” tab of the gexf visualization (on retina.cortext.net)?
Thanks in advance for your reply!
The two “weight” values are coming from two completely different variables.
To find the weight, shown in the Explore” tab of the gexf visualization (on retina.cortext.net), you should open or use the files which are in the folder :
- maps-output\maps\maps_export_format\gexf\ (with gephi, or other gexf reader tool)
In a nutshell:
- “weight” in the clusters.csv, corresponds to the keyword contribution to the cluster (and stream / suprathematique if there are time periods), so basically how central is the term in the cluster. The values of this variable are used to select terms for cluster labels (and stream / suprathematique labels if there are time periods).
- “weight” in the gexf and in the Explore” tab of the gexf visualization (on retina.cortext.net), is related to the weight of the term in the dataset, base on the choice made on the Node weight parameter.
I hope it helps!
Thanks a lot Lionel!
I’ll follow up your answers with two questions:
– Files in maps-output\maps\maps_export_format\gexf\ are the maps I can find in /mapexplorer and open with CorText, isnt-it ?
– Can I find somewhere a table with all the nodes with their weight in the dataset (so not only their weight in their cluster as shown in clusters.csv)?
Just open your network’s maps-output\maps\maps_export_format\gexf\ of your network with Gephi, for example. And :
- then click on the “Data Laboratory” tab ;
- in the datasheet, see the sub tab “Nodes”, or click on it
- If you want to have the data in a csv/excel file, just export the table using “Export the table” button.
I hope it helps,