A Proposed S.C.O.R.E. Evaluation Framework for Large Language Models : Safety, Consensus, Objectivity, Reproducibility and Explainability