Article Preview
TopIntroduction
Analytical tools and models that allow understanding students’ behavior can infer individual or collective patterns and improve students’ experiences (Saito & Watanobe, 2020). These mentioned tools and models can also help teachers monitor the learners’ actions in a Learning Management System (LMS) (Baneres et al., 2019; Weiand et al., 2019).
In this context, new information generated from LMS data can facilitate the teaching and learning process. LMS collects data about users that can help define the learner’s profile, learner’s behavior, and identify their difficulties and needs. One way of accompanying learners is to observe the actions they perform on the system, and these actions can result in paths known as Learning Paths (LP).
The literature works related to LMS consider the LP from two points of view (Ramos et al., 2015):
- 1.
The preferential sequence of access to content (such as educational resources and activities) defined by teachers when planning the course. They made available this sequence in the LMS;
- 2.
The path covered by students during their interactions with the content provided in the LMS.
This work presents a new model of LP based on graphs to assist teachers in their activities. In this work, the term LP refers to the trajectories that students go through. The LP data are obtained based on the student interaction actions with the LMS. The use of stored data in the LMS is already, in many cases, enough to help in decision making, as shown in the experiments. The idea is to propose an alternative for the current user model, which represents the student’s domain knowledge state (Huang et al., 2016). So, this paper’s main contribution is to demonstrate that data that does not reflect, at least directly, the student’s knowledge level, can also be used to create a model of learning paths with the ability to be applied in several different learning environments and to serve as a basis for the creation of new tools.
Monitoring students manually is hard work for teachers, especially when they need to deal with large classes. Thus, based on the proposed model, two tools were created to support the teacher, one to monitor the student’s behavior and the other to assist in the groups formation.
Another contribution of this work is to present a more straightforward approach because, in most works, specialists must analyze and classify resources/activities. They also need to indicate attributes such as level of difficulty, expected time of execution, required knowledge, what students learned in that content, among others, and those actions are not necessary for the proposed approach.
The next sections will present the LP model proposed, a review of the literature, materials and methods, the tools and the experiments carried out with them, and, finally, the discussion of the results and conclusions.
TopLog-Based Learning Path Model
This section shows the used data source and how the proposed LP model was designed based on that data.
Data for Learning Paths
Our LP model uses the student’s actions history from the database records, requiring no more direct data collection, such as filling in forms, surveys, or questionnaires. Most LMSs use resources and activities. Resources are content, in general, static, also called teaching materials. The activities are contents that require the more active participation of the learners, allowing the interaction between them or being a tool of evaluation. Tasks, chats, forums, and quizzes are some examples of activities.
Our approach constructs the paths using cross-data of several tables. The central point is the logs that indicate when and what resources and activities the students accessed. It is possible to chronologically order events that occurred during the student’s interaction with the LMS.
Once ordered, logs present the student’s browsing sequence by the resources and activities of the LMS. From the observation of this structure and previous studies (Ramos et al., 2015; Ramos & Oliveira, 2015), it was concluded that it was possible to use the graph structure to represent the LP.