An alternative m-learning scoring approach to automatically evaluate Iqro’s pronunciation based on pitch, volume, and rhythm features

Abstract

<span lang="EN-US">The </span><em><span lang="EN-US">Iqro’</span></em><span lang="EN-US"> has been proven as an effective method to improve literacy for reading </span><em><span lang="EN-US">Hijaiyah</span></em><span lang="EN-US"> letters which are the fundamental skill for reading the Qur’an. The </span><em><span lang="EN-US">Iqro</span></em><span lang="EN-US"> method relies on a mentor to evaluate the accuracy of the student’s pronunciation of the </span><em><span lang="EN-US">Hijaiyah</span></em><span lang="EN-US"> letters. With the advancement of mobile technology, many mobile apps are available to assist the learning process. However, the existing solution that is available on the market only provides a one-way learning experience where users only focus on enhancing vocabulary, writing, and reading ability with a simple true-false assessment. Although enhancing the vocabulary and reading ability is important, training the correct </span><em><span lang="EN-US">Makhraj</span></em><span lang="EN-US"> (Arabic pronunciation) is also essential in fundamental </span><em><span lang="EN-US">Hijaiyah</span></em><span lang="EN-US"> learning. This paper presents the feasibility investigation of the usage of the pitch, volume, and rhythm as the more comprehensive audio assessment parameters in the voice recognition module to evaluate the student's ability to pronounce the </span><em><span lang="EN-US">Hijaiyah</span></em><span lang="EN-US"> letter in an m-learning implementation. Instead of just presenting the true-false output, the proposed method will calculate the </span><em><span lang="EN-US">Makhraj</span></em><span lang="EN-US"> correctness in the range 0 to 100. Experimental results with 10 respondents show that the sample correlation coefficient using Pearson between a manual assessment by </span><em><span lang="EN-US">Ustaz</span></em><span lang="EN-US"> and the application result for </span><em><span lang="EN-US">Iqro’</span></em><span lang="EN-US"> level 1 is 0.51 which means this approach is acceptable for future implementation.</span>

Bangla song genre recognition using artificial neural network

“…A song is a piece of music created with the intention of being sung by a human voice. This is frequently done at specific, fixed pitches [1], such as melodies that employ sound and silence patterns. Songs come in a variety of formats, such as those with sections that are repeated and varied, which determine the genre of the song.…”

A review on learning analytics in mobile learning and assessment

“…Learning analytics can also be applied in the context of assessments, whereby the data is derived from assessment psychometrics data, with the outcomes used to provide feedback to learners. For instance, the Hijaiyah m-learning application provides a comprehensive audio assessment and users who use the application can read Hijaiyah letters with correct Makhraj pronunciation [9]. In another example, learning analytics is used for predicting students' academic achievement based on "Total login frequency in LMS," "(Ir) regularity of learning interval in LMS," and "total assignments and assessment composite" [10].…”