Enhancing students performance through dynamic personalized learning path using ant colony and item response theory (ACOIRT)

Generating personalized learning paths (PLP) tailored to each student's characteristics and abilities is challenging, especially when using multiple-choice test scores to predict knowledge levels, which can be affected by guessing. We combined the ant colony algorithm (ACO) with Item Response Theory (IRT) to address this issue. This method models the relationship between student responses and their latent abilities, to improve the prediction of knowledge levels resulting in more accurate personalized learning paths (PLP). The parameters used to create the PLP include the initial module, targeted module, knowledge level, pretest score, and difficulty level. The difficulty level is required to ensure the learning object aligns with the student's abilities. Additionally, the system retains a dynamic personalized learning path that accommodates changes in the student's abilities in each module. We divided 80 students into Group X (using ACOPLP) and Group Y (using ACOIRT), each with 40 students. We conducted a pretest and posttest on both groups in order to analyze the effects of implementing the PLP system. Group X outperformed Group Y on the pretest, however, Group Y surpassed Group X on the posttest, with score enhancements ranging from 60.8% to 127.8%. The statistical analysis revealed a noteworthy enhancement in performance for Group Y, with a p-value of 0.002. These results indicate that ACOIRT has the potential for improved student’s performance.