Quantum mechanics

World Wide Web

Scale (ratio)

Engineering

Materials science

Social psychology

Aggregate (composite)

MovieLens

Data mining

Computer science

Social network (sociolinguistics)

Physics

Operations research

Organic chemistry

Collaborative filtering

Psychology

Mathematics

Social media

Political science

Group decision-making

Artificial intelligence

Partition (number theory)

Composite material

Cluster analysis

Combinatorics

Chemistry

Recommender system

Group (periodic table)

Machine learning

Knowledge-Based Systems

With the development of big data and social computing, large-scale group decision making (LGDM) problems attract much attention and become merging with social networks or behavioral factors. In this paper, by considering the trust of social behavioral factor, a two-stage trust network partition algorithm is proposed to reduce the complex of LGDM problems. The large-scale decision makers (DMs) are classified into some leader–follower sub-networks through a network partition algorithm. And a solution method based on trust relationship is proposed to keep the independency of the sub-networks when one follower belongs to more than one leader. Next, weights of independent sub-networks and their individual members are computed, and are further used to aggregate the comprehensive decision information. Finally, alternatives of LGDM problems are sorted with the comprehensive decision information. An experiment with MovieLens data is given to illustrate the proposed LGDM algorithm. And a comparison analysis with general clustering method is provided to verify the effectiveness and feasibility of the proposed method.

A two-stage social trust network partition model for large-scale group decision-making problems