International Journal of Computing and Digital Systems ISSN (2210-142X) Int. J. Com. Dig. Sys.13, No.1 (May-23) http://dx.doi.org/10.12785/ijcds/1301116 A New Spell-Checking Approach Based on the User Profile Ahmed AbdAlrhman Saty 1 , Si Lhoussain Aouragh 2 and Karim Bouzoubaa 3 1 College of CS and IT, Sudan University of Science and Technology, Sudan. 2 High National School for Computer Science and Systems Analysis, Mohammed V University in Robat, Morocco. 3 Mohammadia School of Engineers, Mohammed V University in Robat, Morocco. Received 2 Aug. 2022, Revised 2 May. 2023, Accepted 23 May. 2023, Published 30 May. 2023 Abstract: This paper presents a new approach for spell-checking based on the user profile and that can be applied for any language. For this purpose and for the specific case of Arabic, spelling errors are studied and divided into 18 types. Then, a relationship model between users and their errors is obtained. The proposed architecture initially gives apposite values for a current user, then corrects misspelled words by applying the spelling rules, and the remaining words are corrected based on the probability given by an adopted model of the profile values. To show the efficiency of our profile-based approach, we conducted an experiment with a corpus of 11,908 words containing 1,888 errors. It showed that our approach suggests the correct word in 88.43% times and ranks it in the first four positions in 75.14% times. Moreover, using the same corpus we compared our implemented tool with two existing ones where ours ranked better in 69.79% times than Sahehly and 77.63% times than MS word. Keywords: Arabic linguistic, error-type, intelligent profile detection, multinomial logistic regression, spell-checking 1. INTRODUCTION Spell-checking is based mainly on checking unaccept- able written words in the used language and suggesting a list of correct words that are related or similar to the given incorrect word. The error [1] may be an isolated error, meaning that the written word does not belong to the language in use, or a real-word error, meaning that the written word is correctly spelled but is not used in the correct position [2], [3]. The importance of spell- checking is increasing with the proliferation of computers and smart devices in most aspects of daily life. Spell- checking approaches and algorithms [4], [5], [6], [7] can be adapted to any language, taking into account that a vocabulary and a rule-based module should correspond to the characteristics of the chosen language. In this paper, we focus on the isolated errors in the Arabic language, noting that we use the Buckwalter transliteration [8] convention. Arabic has a rich morphology and a templatic derivation that makes it possible to form many lemmas and words from a given root. Also, it has affixes (prefixes and suffixes) that can be added to a word [9], [10]. Some letters are spelled differently depending on their position in words. For instance, the hamza (“ ”) is written in five different forms depending on its position such as “   ” (“sWAl”), “   ” (“sOl”), and “      ”(“Os}lp”). Moreover, many Ara- bic letters are similar in their shapes, such as “ ” (“E”) and “  ” (“g”), or in pronunciation, such as “  ” (“D”) and “ ” (“d”). Besides, the process of abbreviation in some words, such as the words “   ” (“mn mA”) is abbreviated to “ ” (“mmA”). All these factors and various levels of users make it important to study spell-checking based on a deep understanding of the specificity of the composition of Arabic. As mentioned earlier, Arabic digital content is con- stantly increasing in the Arab and Islamic digital world, and the widespread use of electronic services and social multimedia services such as Facebook and WhatsApp have also contributed to the increase in the number of users. Consequently, with the rapid growth of Arabic digital con- tent and users, as well as the specificity of many linguistic phenomena as explained above, user spelling errors are increasing. Accordingly, it is appropriate to find a new approach for the Arabic Spelling checkers which satisfies and handles all user types, starting with detecting the user profile from the categories, and then choosing the suitable error corrections of the current user. This paper presents a new technique that improves Arabic spell-checking by developing an approach based on the user’s profile. The rest of the paper is organized as follows. Section 2 reviews the related works. Section E-mail address: wdsaty@hotmail.com, l.aouragh@um5r.ac.ma, karim.bouzoubaa@emi.ac.ma http:// journals.uob.edu.bh