Hello, In a researcher’s group, we use Gargantext to analyze 30-40 scientific publications. We mainly import CSV files (source, title, abstract, etc.) or directly with a Gargantext’s Calc.

The termlist records are very different between V3/V4 :

• GarganText V3 = All terms 376 records
• GarganText V4 = All terms 65 records

For small corpora, the termlist V4 is not sufficient to generate a graph. Would it be possible to expand the selection of « candidate terms »?