IEEE Systems, Man and Cybernetics Magazine - April 2023 - 10

process of completing these projects. Next, the remaining
learners need to be assigned to projects on the premise of
ensuring the optimal overall team performance on all
projects. We identify these learners as compatible with a
leader and assign them to a team, with the same assumption
that " learners with a higher degree of compatibility
with the leader will lead to a more satisfactory collaborative
learning outcome. "
In the second type of LC-LTF, conflicts remain not only
among the learners but also between learners and projects.
Learners, whose competence values do not meet the
requirements of a project, are not suitable to participate in
the project even if they have high compatibility values with
the leader. Taking the project in Table 2 as an example, to
complete an exercise system via cognitive diagnosis, learners
who share high compatibility with the leader but lack
the skills related to cognitive diagnosis will not be
assigned to the project. The problem is illustrated in Figure
2. In this problem, it is necessary to calculate the compatibility
between the ability of learners and the
requirements of the projects. Fuzzy interval numbers are
used to represent ability values and project requirements
in this article.
Collaborative Optimization Framework
The preceding statements about the two types of LC-LTF
problems show that LC-LTF problems actually share several
common features in nature. For example, every learner
must play a specific role in a learning team, different
roles have different requirements for the role-players,
learners have different competences to play different
roles, and the global optimization of all teams' learning
outcomes is expected. Just based on them, we believe that
the two types of LC-LTF problems could be essentially
regarded as an RBC problem. Meanwhile, inspired by the
ideas of agent-based computing [26], [27], we model learners
as agents and propose an optimization framework
based on the E-CARGO model, presented in Figure 3. It
includes an evaluation module, a modeling module, and a
solving module.
Evaluation Module
To divide learners into different teams more reasonably
and accurately, the ability of each learner should be
evaluated. As the blue rectangle demonstrates in Figure
3, we could comprehensively model a learner from five
evaluation indexes, such as cognitive ability, learning
style, personality, leadership, and communication skill.
Specifically, cognitive ability reflects a learner's learning
ability. The cognitive ability of a learner could be evaluated
via his or her test scores. Learning style reflects learners'
individual characteristics of learning. As one of the
most popular learning style models, the Felder-Silverman
model is applied more broadly than other models because
of its greater reliability and accuracy [28]. Personality
could provide decision support for potential conflicts
among learners. Currently, the Big Five model of personality
and its scales has also been developed for the
Chinese [22]. Leadership and communication skills measure
the basic competencies of a leader for managing a
team and organizing group activities [29], [30].
Evaluation Methods
There are various methods to evaluate one learner's abilities;
for example, one learner's cognitive abilities could
be obtained via a pretest [20], [31], [32], [33], and questionnaires
can also be utilized to identify learners' learning
styles, leadership skills, communication skills, and
personality [14], [20], [21], [34]. To better model learners'
cognitive abilities, the cognitive diagnosis methods in
psychology [35], [36], [37] can be introduced to achieve
fine-grained modeling of learners' cognitive status about
knowledge concepts. In addition, considering the uncertainty
and fluctuation of learners' abilities, to more comprehensively
and objectively depict learner's abilities,
fuzzy interval numbers are introduced in this article.
Assign
Projects
to be Solved
......
Projects
1
2
n
Learners
......
1
[[0.8, 1.0], [0.6, 1.0], ...]
Expected Proficiency
[[0.6, 1.0], [0.9, 1.0], ...]
.........
[[0.4, 1.0], [0.6, 1.0], ...]
Learners
2
[[0.4, 0.6], [0.4, 0.5], ...]
Proficiency
[[0.5, 0.7], [0.7, 0.9], ...]
.........
m [[0.2, 0.3], [0.8, 1.0], ...]
Assign
Figure 2. LC-LTF for project-based learning.
10 IEEE SYSTEMS, MAN, & CYBERNETICS MAGAZINE April 2023
Project 1 Learner 2
.........
Learner 1 Learner 2
.........
Learner-Project Conflicts
.........
Member Leader Member
Member Leader Member
Learning Team 2
Learner-Learner Conflicts
Learning Team 1
Learners
Evaluation
Projects
Evaluation
Constraints
Formulation
Team
Formation
IEEE Systems, Man and Cybernetics Magazine - April 2023

Table of Contents for the Digital Edition of IEEE Systems, Man and Cybernetics Magazine - April 2023
