Rankings of different research bodies are of particular interest for academia and bibliometrics is used to measure the quality of these research bodies. Different factors affecting this quality have been proposed. In this paper, we have demonstrated a new approach based on fuzzy models and taken into account different proposed factors to access the overall quality. Our model actually refines the ranking by adding more factors which affects this quality. Thus, improving the information retrieval system based on human reasoning. A dimensionless index called Fuzzy Index (FI) has been proposed and used to shuffle the previously ranked research bodies. We have successfully demonstrated the application of our FI in ranking journals, conferences and authors in the field of computer graphics in particular and computer science in general.