Graduate School of Fundamental Science and Engineering Waseda University
博 士 論 文 概 要
Doctoral Thesis Synopsis
論 文 題 目
Thesis Theme
Effectiveness of Reversing the Image of
Roman Alphabet-Based Transcription on the Acquisition of Arabic in E-Learning
e ラーニングでの アラビア語習得に対する 反転欧字発音表記の効用
申 請 者 (Applicant Name)
Ahmed Ebrahim Mohamed MOSA モサ エブラフィム モハメド アハメド
Department of Computer Science and Engineering Research on Software Environment
December, 2015
E-learning is important for learning languages. Transcription is important support for reading/pronouncing foreign languages. Arabic language is one of the popular languages all over the world. There are many reasons that make learning Arabic difficult for many foreigners; Arabic script is one of these reasons. There are 29 letters and 10 diacritical marks. A sentence is a sequence of words arranged from right-to-left. A word is a sequence of letters with attached diacritical marks arranged from right-to-left. In Arabic, most letters forming a word are connected to each other in a word. Among the 29 letters, there are 22 letters have four forms. One form is independent and three are dependent. Dependent form is selected according to its position in a word. And 7 letters have only an independent form. These letters that can be connected to the preceding letter but cannot be connected to the following letter. There are three vowels “a”, “i” and “u” used in Arabic language. An Arabic letter and its attached diacritical mark present a syllable.
In current e-learning systems of Arabic Acquisition, Roman Alphabet-Based Transcription is one of the most common reading supports. Roman transcription is arranged left-to-right. Learners do not pay attention to read or recognize Arabic script. They can easily read the transcription itself. Moreover, the number of Roman letters in the transcription is not equal to the number of Arabic syllables in a word. In addition to, both Arabic letters and their attached diacritical marks are transcribed by lowercase letters. For learning Arabic language, our target is supporting the non-Arabic speaking beginners to read Arabic syllables in words from right-to-left. For that, current Roman transcription in e-learning does not work well and learners find difficulty to identify and read Arabic syllables in words.
It is important to provide a transcription, from right-to-left, having a property of finding out the letter(s) that is corresponding to every Arabic syllable in a word in e-learning.
In this thesis, we show the developing and effectiveness of reversing the image of Roman Alphabet-Based Transcription in an e-learning system on supporting the beginners in learning the Arabic language especially identifying and reading Arabic syllables in words.
We invented a new transcription for Arabic script using Roman letters. Our transcription consists of units; a unit transcribing an Arabic syllable consists of two Roman letters: one in uppercase that corresponds to the Arabic letter, and the other in lowercase that corresponds to the attached diacritical mark. We arranged the units from right-to-left.
And then we reversed the image of all Roman units horizontally to match the direction of Arabic script. We call our Reversed Image of right-to-left Transcription “RIT”. We developed a new version of RIT. We call it RIT’, by which each Arabic syllable is transcribed by Roman unit, but each unit consists of a mixture of a reversed uppercase Roman alphabet and a copy of the diacritical mark of that Arabic syllable.
No 1
We determined how to support the beginners in identifying Arabic syllables in words. We conducted a user study to know which one is a better support for the learners, coloring or separating Arabic syllables in words.
The results showed that, among 77 learners, 60 preferred coloring the Arabic syllables more than separating them. The learners were categorized on male/female, interest/no interest in Arabic and have/have no experience in Arabic. There were no significant differences between the percentages of preferring coloring syllables on the two attribute values in those three categories.
We decided to show the syllables of Arabic script in two different colors for supporting the beginners to identify the syllables of Arabic in words. And then we decided to show the units of RIT and RIT’ in two different colors corresponding to the Arabic syllables to support the beginners to find out the transcription unit of each Arabic syllable in same direction and same color. We call colored RIT and RIT’ as RITc and RIT’c.
We investigated the effectiveness of RIT on supporting some examinees in starting the reading from right-to-left comparing to other two ways of transcriptions. First transcription is arranging all Roman letters from right-to-left. Second transcription is arranging all Roman units from right-to-left. Through a workshop, the examinees read RIT from right-to-left. However, most of the examinees read the other two transcriptions left-to-right.
We investigated the effectiveness of RIT compared to one of the conventional transcription called BATR.
This is on supporting the learners in reading different Arabic syllables located in different words. We conducted a workshop and a questionnaire with 34 learners having no experience in Arabic in that time. The results showed that almost all of the learners used RIT correctly to identify the units in the Roman transcription for the Arabic syllables at any position in a feasible time. They also read the located units correctly. All the learners preferred RIT to BATR in reading Arabic syllables in script. Also, there were 8 volunteers having experience in Arabic and they preferred RIT to BATR.
We investigated the effectiveness of RITc and RIT’c in computer-based learning on supporting the beginners in learning Arabic through Arabic course of six Arabic lessons. The supports are colored Arabic syllables, RITc corresponding to Arabic syllables in first two lessons and RIT’c corresponding to Arabic syllables in next three lessons. Support of the last lesson is showing a table including RIT for all Arabic letters and diacritical marks.
The system provides supports by which learners can read explanation, listen to correct reading, watch video club and illustration for learning the diacritical marks and the Arabic specific sounds. We conducted a workshop and a questionnaire with 59 beginner learners. The learners had a variety of nationality, age and gender. We asked each learner to study the diacritical marks, Arabic specific sounds and the Arabic lessons one by one by the computer-based learning. There was a possibility for each learner to discuss with a teacher of Arabic and to get feedback on his/her progress in each lesson.
No 2
The results showed that RITc and RIT’c in the computer-based learning supported the learners in studying Arabic well. Almost all of learners succeeded in reading more than 70% of the Arabic syllables, recalling and reading more than 80% of the Arabic expressions, and reading and recognizing more than 90% of the Arabic expressions with and/or diacritical marks. Moreover, average of learning time has been reduced from 142 to 87 minutes through the six lessons. Average of learning as a whole is around 13 hours. Arabic teachers reported that this progress of learning takes from 16 to 24 hours. Some of the learners reported that RITc is helpful and most of them reported that RITc is strongly helpful for them in reading Arabic. Most of the learners welcomed RIT’c.
We investigated the effectiveness of RITc and RIT’c in e-learning system prepared by MOODLE and called ETaJWa of five Arabic lessons. This is on helping the learners in reading Arabic syllables and expressions, identifying talking to male and female and identifying talking about male and female. There were 14 learners covering variety of the general characteristics who learned the Arabic lessons and there were other 22 volunteers who just explored ETaJWa to answer a questionnaire of 16 benefits of ETaJWa and its multimedia. The learners were of different abilities in reading Arabic.
The results showed that RITc and RIT’c in ETaJWa supported the learners in reading more than 75% of the Arabic syllables, recalling and recognizing more than 79% of the Arabic expressions, and reading more than 81% of the Arabic expressions with diacritical marks. Among 14 learners, 13 who identified correctly the talking to male and female and identified two pronouns for “he” and “she”. Twelve who identified two demonstrative pronouns for “this” as male and female. Moreover, almost all of the learners could write their names correctly as activities without help.
The results of questionnaire before the learning showed that 12 among 14 learners agreed that RITc is helpful for them, as a support compared to the available left-to-right transcriptions, to read Arabic syllables. The others agreed that RITc is strongly helpful. 19 among 22 volunteers agreed that RITc is a helpful for them to read Arabic syllables. The other two learners agreed that RITc is a strongly helpful. The results showed also that 8 among those 14 learners agreed that RITc is helpful for them, as a support compared to the listening to a teacher pronunciation support, to read Arabic syllables. The other 4 learners agreed that RITc is a strongly helpful. 14 among 22 volunteers agreed that RITc is a helpful for them to read Arabic. The other 6 volunteers agreed that RITc is a strongly helpful.
The results of another questionnaire after the learning, regarding 16 benefits of ETaJWa and its supports, showed that most of the learners and volunteers agreed that the supports are strongly helpful and the others agreed that the supports are helpful in learning Arabic.
We conclude that applying the reversed image of the Roman alphabet-based transcription corresponding to Arabic script in the e-learning ETaJWa, is a helpful support for the non-Arabic speakers who are adults to study Arabic language especially start reading Arabic script. Our supports in ETaJWa do not lure the learners to stay in the transcription and unconsciously encourages them to read out Arabic script directly without help. So that, the reversed image transcription is a good support for Arabic acquisition.
No 3
No.1
早稲田大学 博士 エ 学 学位申請 研究業績書
(List of research achievements for application of doctorate Dr. of Engineering, Waseda University
氏 名(Full Name) Ahmed Ebrahim Mohamed MOSA 印(seal or signature )
（As of December, 2015） 種 類 別
(By Type)
題名、 発表・発行掲載誌名、 発表・発行年月、 連名者（申請者含む）
(theme, journal name, date & year of publication, name of authors inc. yourself) Journal
Journal
International conference
International conference
Presentation
International workshop
A Way of Supporting Non-Arabic Speakers in Identifying Arabic Letters and Reading Arabic Script in an E-Learning System. Research Article in EAI Endorsed Transactions on e-Learning, 15(6): e3. (2015). Mosa, A. and Kakehi, K.
Effectiveness of our New Right-to-Left English Transcription on Arabic Learning Using the Reversed Image -Support in Arabic e-Learning System for the Beginners- Global Journal of Computer Science and Technology. GJCST. Volume 15 Issue 4 Version 1.0, pp. 35-45, (2015).
Mosa, A. and Kakehi, K.
Searching for a Suitable Way to Transliterate Arabic into Roman Letters as for a Device in Arabic e-Learning Systems. International Conference: e_Society 2013: IADIAS. Lisbon, Portugal, March 13-16, 2013. Proceedings Page Number 448-452. (2013). Mosa, A. and Kakehi, K.
Letting Non-Arabic Speakers Read and Pronounce Arabic Sounds Using Roman Phonetic Codes in the Mirrored Form -a Feature of ETaJWa, an Arabic e-Learning System-. In: 2nd International Conference on Computing, E-Learning & Emerging Technologies (ICCEET 2014), Paris, France (July 8-9, 2014), pp. 59-66. (2014). Mosa, A. and Kakehi, K.
Towards a Multimedia Based e-Learning System for Effective Acquisition of Arabic language.
IPSJ SIG Technical Report (CLE), Japan, 2012- CLE- 7(3), 1-3. (2012). Mosa, A. and Kakehi, K.
Designing a New Transliteration System of Arabic Language into Roman Letters as for a Device in Arabic e-Learning Systems. Seoul, Korea. 2013. 11. 8 (Fri.):11. 9 (Sat.). Hanyang- Waseda IT WORKSHOP. Hanyang University. (2013). Mosa, A. and Kakehi, K.
