Evaluation Mechanism of Collective Intelligence for Heterogeneous Agents Group

Abstract

Collective intelligence is manifested when multiple agents coherently work in observation, interaction, decision-making and action. In this paper, we define and quantify the intelligence level of heterogeneous agents group with the improved Anytime Universal Intelligence Test(AUIT), based on an extension of the existing evaluation of homogeneous agents group. The relationship of intelligence level with agents composition, group size, spatial complexity and testing time is analyzed. The intelligence level of heterogeneous agents groups is compared with the homogeneous ones to analyze the effects of heterogeneity on collective intelligence. Our work will help to understand the essence of collective intelligence more deeply and reveal the effect of various key factors on group intelligence level.

Evolutionary emergence of collective intelligence in large groups of students

“…The analysis of top-down factors has focused on the interaction among the members of a group, analyzing several aspects such as team creativity, group heterogeneity, individual incentives, consensus-seeking, duration and continuity of the interaction, and the successive order of turns taken by the group’s members ( De Vincenzo et al, 2017 ; Aggarwal and Woolley, 2018 ; Bernstein et al, 2018 ; Dai et al, 2020 ). This line of research tends to focus on situations with small groups (between 2 and 5 individuals), which makes it difficult to study how collective intelligence behaves according to group size, particularly in the case of large groups.…”

“… Mao et al (2016) point out the need of investigating the behavior of such large groups in experiments featuring a limited number of variables, and taking place in natural contexts that reflects the complexity of real-life situations. Analyses have thus been conducted on simulations that study a group’s degree of heterogeneity ( Dai et al, 2020 ), as well as a group’s tendency to consensus ( De Vincenzo et al, 2017 ; Massari et al, 2019 ) the social learning and the group size ( Garg et al, 2022 ). In experimental situations and natural contexts, Tinati et al (2014) studied the problem of collaborative creation in their analysis of citizen science projects and the different levels of involvement of volunteers.…”

Interaction among participants in a collective intelligence experiment: an emotional approach

“…Among possible bottom-up factors, interaction among group members is regarded as a necessary condition for collective intelligence to emerge ( De Vincenzo et al, 2017 ; Bernstein et al, 2018 ; Dai et al, 2020 ). Indeed, within the paradigm established by Woolley et al (2010) , the condition of interaction is essential, as the final product is achieved by a group.…”