Xianzhong Zhou

Group Role Assignment With Cooperation and Conflict Factors

Collaboration is complex. To solve a problem occurring in collaboration, using computers, we must first define and specify the problem. This paper presents a challenging problem in collaboration, called group role assignment with cooperation and conflict factors (GRACCFs). This problem’s solution aims at creating a high-performance group by role assignment with consideration of cooperation and conflicts between agents. The contribution of this paper is the formalization of the proposed problem, a confirmation of the complexity of the problem, a practical solution that uses the IBM ILOG CPLEX optimization package (ILOG), a verification of the benefits of solving the GRACCF problem by simulations and a practical way of collecting the required factors to support decision makers in solving such a problem within a real-world scenario. Experiments are used to verify the efficiency of the proposed ILOG solution.

Effective Approaches to Adaptive Collaboration via Dynamic Role Assignment

Adaptive collaboration (AC) is essential for group performance optimization in collaborative systems. This paper begins by introducing AC within the context of solving a real-world problem. Next, AC problems are formalized based on the environment-class, agent, role, group, and object (E-CARGO) model. Three algorithms are proposed for solving AC problems. They are based on three different scenarios: 1) the current group state (GS); 2) the GS after a specific period; and 3) the GS throughout the collaboration. More complex AC problems and their solutions are then investigated. Derived from the above-mentioned algorithms, two additional algorithms are presented. They consider reassignment costs. Experiments are developed to analyze the performance of each proposed algorithm. Results indicate that the proposed algorithms perform better than static collaboration where there are no reassignment costs. If reassignment costs exist, we also provide a way of determining whether to adopt an AC approach. This paper provides insights into the AC process and its effectiveness in various scenarios.

Humanware Model in Novel Decision System

How to integrate human (expert) into the procedure of decision-making is an important issue for the next generation of decision support systems (DSS). Based on the concepts related to humanware (Hw) in our research, including Hw, Hw technology, HW service, and Hw and knowledge, the framework of novel decision systems (NDS) is presented. Then this paper proposes the structure of Hw model (HwM), which describes the key component and implement steps in detail. HwM can convert human (expert) into one component, so as to facilitated Hw to participate in the procedure of decision-making. Finally, a case study demonstrates that Hw can be engaged in the procedure of decision-making.

Effective approaches to group role assignment with a flexible formation

Group role assignment with a flexible formation (GRAFF) is essential for group performance optimization in collaborative systems. In this paper, problems of GRAFF are formalized based on the Environment-Class, Agent, Role, Group, and Object (E-CARGO) model. Then, based on group role assignment (GRA) and linear programming (LP), two algorithms are proposed. Experiments are developed to analyze the performance of each proposed algorithm. Results indicate that the proposed algorithms are effective and the linear programming-based algorithm implemented with the IBM ILOG CPLEX package is the most efficient way to solve problems of GRAFF.

A Quantum-inspired Q-learning Algorithm for Indoor Robot Navigation

A quantum-inspired Q-learning (QIQL) algorithm is proposed for indoor robot navigation control. Q- learning is an action-dependent reinforcement learning method and has been widely used in robot navigation. Inspired by the fundamental characteristics of quantum computation, e.g. state superposition principle and quantum parallel computation, probability is introduced to Q-learning and along with the learning process the probability of each action to be selected at a certain state is updated, which leads to a natural exploration strategy instead of a pointed one with configured parameters. The simulated navigation experiments show that the proposed QIQL algorithm keeps a good balance of exploration and exploitation, which can avoid the local optimal policies and accelerate the learning process as well.

Group role assignment based on multiple criteria in collaboration systems

In the real-world, there are many criteria to evaluate agents on roles, which are an important factor affecting the result of the group role assignment (GRA). To obtain a satisfied assignment, we need to select an appropriate evaluation method. This paper points out the key issues of the related concepts, formalizes the problem of GRA based on multiple criteria (GRA-MC), investigates three algorithms of multi-criteria decision making, i.e., simple additive weighting (SAW), multiplication exponent weighting (MEW) and weighted distance (WD), briefs the procedure of GRA-MC, verifies the GRA-MA methods by simulations and a case study, and finally discusses the result of GRA-MC.

Path planning for multi-robot systems using PSO and Critical Path Schedule Method

In recent years, there are more and more situations that need robots to do dangerous and complex works instead of human, to improve the efficiency and reduce the risk. Path planning for multi-robot systems is one of the most important problems to solve in practice. As the improvements in communication and control technology of robots, an algorithm is proposed using Particle Swarm Optimization and Critical Path Schedule Method to do path planning for multi-robot systems to reach the targets without collision not only between the robots but also between the robots and environment. Different from traditional method, a schedule is proposed on the planned optimal or feasible path to enable the robots to avoid collisions and minimize the cost such as time, distance. Moreover, a simulation experiment is given to verify the validity of proposed algorithm, whereby the environment is studied in a two dimensional free space.

Multi-task Assignment in Information Technology Companies Based on the E-CARGO Model

In the process of Research and Development (R&D), the staff in Information Technology (IT) companies forms a collaboration system. Task assignment is very important in the process of R&D. Firstly, this paper formalizes the staff and task assignment problems (STAPs) based on the Environment, Class, Agent, Role, Group and Object (E-CARGO) model. Next, we propose an algorithm for the STAPs. Then, we verify the proposed algorithm with experiments. The experimental result shows that the proposed algorithm is effective.

The benefits of conflict avoidance in collaboration

Conflict management is an important and challenging problem in collaboration. Conflict avoidance is one way to manage conflicts. Intuitively, we may try the best to avoid conflicts when initiating a collaborative activity, because we may save the effort of conflict resolution. However, conflict avoidance is a complex task that requires many efforts. We need to know if we could get a reward from our payment in conflict avoidance. This paper sets up a model to evaluate the collaboration performance before and after conflict avoidance; conducts experiments with simulations; finally conclude the benefits for conflict avoidance. The contribution of this paper is to assert that conflict avoidance is beneficial in most cases unless the conflicts are not severe, there are not many conflicts, and the number of roles is too small.

Salient locations search based on human visual attention: An experimental analysis

To explore whether the conscious options or eye movements of human beings are influenced by bottom-up saliency, both theoretical and experimental analysis are presented in this paper. In the first experiment, the subjects are required to indicate most interesting regions of test pictures in the scene. In the second experiment, the subjects are required to look around test pictures in the scene. Then, the bottom-up saliency values of test pictures are computed by Itti model of that is one well-known visual attention. It is indicated that the interest points and fixation locations are correlated with the bottom-up saliency locations. Thus we propose two methods to study on visual attention that can be used to predict which areas of the images will draw subjects visual attention: fixation location predictions and interest point predictions.