A Proposal for Making Argumentation Computationally Capable of Handling Large Repositories of Uncertain Data