CHAPTER 2
LITERATURE REVIEW
This chapter provides the background for the study that acquired by reading articles, conference paper and journal which will cover the tajweed and speech recognition
2.1 Tajweed
This section will explain about the introduction of tajweed, the type of tajweed that have in the Al-Quran and the reason of learning tajweed using voice.
2.1.1 Introduction
Arabic is the fifth most usually spoken language that makes it one of the world’s major languages (Yousef, 2013). One of the Arabic languages is the Al-Quran. Al-Quran recitation has a rule for reading it which is known as tajweed to make sure proper pronunciation of reading it (Ahsiah&Idris, 2013). The Quranic Arabic alphabets consist of 28 letters which known…show more content… et. al., (2013)
2.1.3 Learning Using Voice
With the voice recognition students or learners can learn by themselves not include expert because they can revise it themselves by hearing back their recitation as mention by Ahsiah (2013) it can help students to learn or revise their correct recitation by themselves. Research being done by Irfan (2015) it is recommended that voice recognition should be embedded in any language learning because it can improve their skill and the learner can listen back to their speech and correct it.
2.1.4 Summary
In conclusion, tajweed has many rule to and need to learn it with step by step because it is the most important things to learn when it is come to recite the Al-Quran and by learning it using voice recognition make the learning become more efficient and fast to adapt.
2.2 Speech Recognition (SR)
This section will explain about introduction of SR, data collection, data pre-processing which is sampling and de-noising, data extraction, classification algorithm for example multilayer perceptron (MLP) and hidden markov model (HMM).
2.2.1…show more content… It is a process of processing a raw data into a meaningful data which includes a sampling and de-noising. First process of data pre-processing method is converts the sound wave captured by microphone into an electrical signals before being extract (Saini, 2013). Data processing being done because the classification cannot extract the data with a raw signal that not be digitalize. Data pre-processing that be held by Posner (2012) sample rate at 8000 Hz in mono channel configuration with a 16 bit encoding format, signal being framed into 20msec segment and signal with higher frequency need to emphasize before processing. Data preprocessing also process the sound so the signal will free from noise that can effects the outcome of the recognition. Pre-processing also consist of filtering processing to remove noise so it does not have an effect on the recognition process (Ahsiah,