What options does Rapidminer offer to create a network graph of individual terms mined from an Excel file of comments? Secondly, I would like to overlay a clustered view on top of these individuals terms that groups like categories?
for such a term graph you could use frequent itemset mining together with association rules - which would deliver the graph. Never tried that on texts myself and I am quite positive, that you have to integrate some appropriate preprocessing. Alternatively, you could maybe use the TransitionGraphOperator.
If you want to overlay a cluster / label / whatever... information, you would have to extend both approaches by this information.
for such a term graph you could use frequent itemset mining together with association rules - which would deliver the graph. Never tried that on texts myself and I am quite positive, that you have to integrate some appropriate preprocessing. Alternatively, you could maybe use the TransitionGraphOperator.
If you want to overlay a cluster / label / whatever... information, you would have to extend both approaches by this information.
Cheers,
Ingo