Forecasting the productivity of a virtual enterprise by agent-based fuzzy collaborative intelligence—With Facebook as an example

Abstract

Since the Internet bubble, firms that focus on virtual enterprises have sought to enhance productivity. To achieve this goal, a firm must evaluate its present productivity and estimate its future productivity. To overcome the considerable uncertainty in estimates of productivity, we propose an agent-based fuzzy collaborative intelligence approach that predicts productivity. First, a fuzzy learning model is built and used to estimate future productivity. Subsequently, the fuzzy learning model is fitted by several agents with diverse settings; those agents produce different productivity forecasts. Fuzzy intersection is then applied to determine the narrowest range that contains the actual value from the fuzzy forecasts. Finally, a back-propagation network derives a representative value from the fuzzy forecasts. The real-world case of Facebook is used to demonstrate the applicability of the proposed methodology.