Abstract:

It is almost inevitable that during a course, students will get distracted either by internal stimuli (e.g., thoughts, attempts to retrieve information from memory) or external stimuli from the physical (e.g., visuals, sounds) or digital environment (e.g., applications). Literature confirms that introducing engaging activities into the main lecture and changing pedagogies within a class period has remarkable effects on students' concentration. However, educators are not always willing to adopt and apply active learning techniques in their teaching routine, since not only the available class time is limited but also the preparation of the required material is a time-consuming task. In a “smart classroom” though, where the computational resources surplus, an intelligent decision-making system could automatically compile and initiate the appropriate activities based on the context of use, the students’ needs and the available time. Therefore, in such environments, a mechanism that monitors the learners and intervenes to reset attention levels can be employed to promote learning.

LECTOR is a framework that aims to take advantage of the ambient facilities that “smart classrooms” have to offer and enable educators to employ their preferred attention monitoring strategies (including any well-established activity recognition techniques) in order to identify inattentive behaviors. Furthermore, LECTOR assists the educators in reengaging the students to the educational process by intervening, when necessary, to (i) provide a motivating activity to a distracted student or (ii) suggest an alternative pedagogical approach that would be beneficial for the entire classroom (e.g., a different lecture format). This paper presents LECTORstudio, an educator-friendly authoring tool for tailoring the logic of LECTOR’s decision making components in order to satisfy the needs of the actual orchestrators of the educational process, the teachers.

It enables the creation or modification of:
-The rules that signify inattention. An intuitive UI allows the definition of the conditions under which the student is considered distracted (e.g., talking during an exam is an indication of inattention) through combining any of the available parameters (e.g., identified activity, context, student profile, past activity). Furthermore, considering that students have diverse backgrounds, experiences and prior knowledge, the educators can create rules personalized to specific learners, by taking into account individual cognitive styles, learning preferences, needs and interests of each individual.
-Class-wide or personalized interventions. The educators are able to create their own interventions or customize the content of any of the existing ones. When creating an intervention, the educator defines whether it can be applied to all the students at once (i.e., class-wide) or it should be presented individually to the distracted students only (i.e., personalized).
-The rules that indicate the conditions under which an intervention is initiated (e.g., percentage of distracted students, type of lecture).

A heuristic evaluation experiment was conducted to eliminate any major usability errors before proceeding with user testing. A small group of five evaluators inspected the interface and judged its compliance with recognized usability principles ("heuristics"). The next steps include the realization of a full-scale user-based evaluation with the targeted end-users.

Keywords:

Authoring tool, attention rules, intervention management, smart classroom.