Meta measurement of intelligence with crowd network

Abstract

Purpose: The multiple factors of intelligence measurement are critical in intelligent science. The intelligence measurement is typically built as a model based on multiple factors. The different agent is generally difficult to measure because of the uncertainty between multiple factors. The purpose of this paper is to solve the problem of uncertainty between multiple factors and propose an effective method for universal intelligence measurement for the different agents. Design/methodology/approach: In this paper, the authors propose a universal intelligence measurement method based on meta-analysis for crowd network. First, the authors get study data through keywords in the database and delete the low-quality data. Second, they compute the effect value by odds ratio, relative risk and risk difference. Then, they test the homogeneity by Q-test and analyze the bias by funnel plots. Third, they select the fixed effect and random effect as a statistical model. Finally, through the meta-analysis of time, complexity and reward, the weight of each factor in the intelligence measurement is obtained and then the meta measurement model is constructed. Findings: This paper studies the relationship among time, complexity and reward through meta-analysis and effectively combines the measurement of heterogeneous agents such as human, machine, enterprise, government and institution. Originality/value: This paper provides a universal intelligence measurement model for crowd network. And it can provide a theoretical basis for the research of crowd science.