Youness CHAABI - Academia.edu (original) (raw)

Papers by Youness CHAABI

International Journal of Emerging Technologies in Learning (iJET)

The present article is elaborated in the context of e-learning softwares that provide assistance ... more The present article is elaborated in the context of e-learning softwares that provide assistance and functionalities to learners engaged in distance learning. Our contribution consists of a system that estimates a behavioral (sociological) profile for each student. This estimation is based on automatic analysis of students’ textual asynchronous conversations. In general, the automatic analysis of textual conversations is based on speech acts for classification and categorization of messages. This technique has several disadvantages like the absence of standardization of speech acts for determining social behaviors of learners. To overcome this, we propose a multi-agent system based on fuzzy logic reasoning and supervised learning technique for automatic classification and categorization of textual conversation. The determined profiles are proposed to teachers to provide them assistance during tutoring tasks. The objective of this article is to share our reflections around these issu...

2017 International Conference on Advanced Technologies for Signal and Image Processing (ATSIP), 2017

This paper proposes a facial expression recognition method based on a novel facial decomposition.... more This paper proposes a facial expression recognition method based on a novel facial decomposition. First, seven regions of interest (ROI), representing the main components of face (left eyebrow, right eyebrow, left eye, right eye, between eyebrows, nose and mouth), are extracted using facial landmarks detected by IntraFace algorithm. Then, different local descriptors, such as LBP, CLBP, LTP and Dynamic LTP, are used to extract features. Finally, feature vector, representing face image, is fed into a multiclass support vector machine to achieve the recognition task. Experimental results on two public datasets show that the proposed method outperforms state of the art methods based on other facial decompositions.

2021 International Conference of Modern Trends in Information and Communication Technology Industry (MTICTI), 2021

Persistence in online courses remains a concern for various institutions. The case of (Massive Op... more Persistence in online courses remains a concern for various institutions. The case of (Massive Open Online Courses) MOOCs represents a particular situation of online courses with an even higher dropout rate. This type of training highlights problems that have already been identified such as sociological isolation of the learner, loss of motivation, empowerment of the learner, acquisition of identity within a group and appreciation of the group pedagogical progress. In this context, the importance of a follow-up by a human tutor is unanimously recognized by the different a ctors. Despite the numerous services offered by open and distance learning platforms, one of the main difficulties encountered by tutors in this task is to have a sufficient understanding of what the distant learners are doing. One way that seems particularly promising to solve this problem is the exploitation of interaction traces left by learners within MOOCs, and the elaboration of indicators that can help the tutor in monitoring learners’ activities. This is why it seems imperative to me to propose a tool to visualize the work accomplished by each learner. The system must provide indicators that help the tutor to appreciate the work of the learners.

Dans cet article, nous proposons un systeme de recommandation de cours en ligne dans une platefor... more Dans cet article, nous proposons un systeme de recommandation de cours en ligne dans une plateforme de formation a distance (LMS), qui assiste et aide les apprenants lors des tests de connaissance. La plupart des LMS commerciaux ou logiciels libres et a code source ouvert n'integre pas d’outil permettant d’envoyer des recommandations afin de guider les apprenants dans leur apprentissage. Les dispositifs de formation en ligne actuels, par exemple les MOOC, enrolent des milliers d’apprenants, ainsi les tuteurs se sentent souvent demunis pour construire une representation synthetique de l’activite des apprenants, d’envoyer des recommandations, des remarques a chaque apprenant et intervenir dans de bons moments. Pour tenter de repondre a ces besoins, nous proposons une approche qui vise a analyser automatiquement les reponses des apprenants qui determinent lors niveau de connaissance. Sur la base de ces reponses, notre systeme propose automatiquement des recommandations personnalis...

There is no doubt that MOOCs constitute a massive phenomenon that has become depressing due to th... more There is no doubt that MOOCs constitute a massive phenomenon that has become depressing due to the skepticism of the traditional stream of distance learning and the reluctance of teachers, practitioners and university pedagogues (Grandbastie, 2018, p. 8). But can we say that the concrete object MOOC is a scientific object in the sense of (Davallon, 2004, p. 8), or is it still only an empirical construction covering realities so diverse that they can not be treated as an only research object? In this perspective, our article sets out to draw up a state of the art of the MOOC horizon and to present a diagnosis of what exists in Moroccan universities based on an empirical work and eclectic methodology.

Most commercial or open source Course Management Systems CMS software does not include comprehens... more Most commercial or open source Course Management Systems CMS software does not include comprehensive access tracking and log analysis capabilities and lack the support for many aspects specific to evaluating participation level and analyzing interactions. CMS does not provide any tools for visually representing ongoing interactions. It is difficult and time consuming for the teachers and the educationalists to ascertain the number of participants, non-participants and lurkers in an ongoing discussion. This paper presents MVSA to Tracking student activity in online course management systems a system that takes a novel approach of using Web log data generated by course management systems (CMS) to help instructors become aware of what is happening in distance learning classes. Specifically, techniques from information visualization are employed to graphically render complex multidimensional student tracking data provided by the Course Management System. MVSA system provides accurate tr...

Dans un contexte presentiel, un acte de communication comprend des expressions orales et emotionn... more Dans un contexte presentiel, un acte de communication comprend des expressions orales et emotionnelles. A partir de l’observation, du diagnostic et de l’identification de l’etat emotionnel d’un individu, son interlocuteur pourra entreprendre des actions qui influenceraient la qualite de la communication. A cet egard, nous pensons ameliorer la maniere dont les individus percoivent leurs echanges en proposant d’enrichir la CEMO (communication ecrite mediatisee par ordinateur) par des emotions ressenties par les collaborateurs. Pour ce faire, nous proposons d’integrer un systeme de reconnaissance, en temps reel, des emotions (joie, peur, surprise, colere, degout, tristesse, neutralite) a la plateforme pedagogique “Moodle” pour les extraire a partir de l’analyse des expressions faciales de l’apprenant a distance lors des activites collaboratives

In this article, we present a recommending system of educational resources based on the footprint... more In this article, we present a recommending system of educational resources based on the footprint of learners on a distance-learning platform. Today, all learners follow the same learning path and therefore have access to the same training content, although their information needs are different. The proposed approach offers learners the opportunity to adapt their training path through a mechanism of recommended activities that are most relevant to them. This approach consists in automatically analyzing the interaction footprints of learners. Based on these footprints, our system automatically offers personalized recommendations (courses, course sections, activities, documents, etc.) to learners. The originality of this approach lies in the fact that the learner remains in control of his training itinerary, by increasing the information available to help him or her to build his or her training path. In this article, we present the results of the implementation of our system developed...

Journal of Big Data, 2019

The authors note a correction to the article [1]. Table 5 of the original article is incomplete. ... more The authors note a correction to the article [1]. Table 5 of the original article is incomplete. Few percentage values are missing. This article presents the corrected version of Table 5.

International Journal of Emerging Technologies in Learning (iJET), 2019

In this article we describe a new multi-agent approach for the accompaniment and follow-up of lea... more In this article we describe a new multi-agent approach for the accompaniment and follow-up of learners (tutoring) in collaborative social networks via network technologies. To assist learners in their collaborative learning process, the system we propose offers the possibility to identify the sociological behavioral’ profile of each learner on the basis of the automatic analysis of the asynchronous textual conversations exchanged between learners. To achieve our aims, we first describe the sociological profiles that we use in our model. Then, we expose the approach used for the semantic analysis of the messages exchanged (full text), as well as the proposed indicators for the determination of these profiles. After, we present the results of the implementation of the system developed as part of an experiment that we conducted with the students of the Master Program “Software Quality” in the Ibn Tofail University of Kenitra, Morocco. We did indeed obtain very good performances during ...

Journal of Big Data, 2018

Gathering public opinion by analyzing big social data has attracted wide attention due to its int... more Gathering public opinion by analyzing big social data has attracted wide attention due to its interactive and real time nature. For this, recent studies have relied on both social media and sentiment analysis in order to accompany big events by tracking people's behavior. In this paper, we propose an adaptable sentiment analysis approach that analyzes social media posts and extracts user's opinion in real-time. The proposed approach consists of first constructing a dynamic dictionary of words' polarity based on a selected set of hashtags related to a given topic, then, classifying the tweets under several classes by introducing new features that strongly fine-tune the polarity degree of a post. To validate our approach, we classified the tweets related to the 2016 US election. The results of prototype tests have performed a good accuracy in detecting positive and negative classes and their sub-classes.

Signal Processing: Image Communication, 2017

Automatic facial expression analysis is a challenging topic in computer vision due to its complex... more Automatic facial expression analysis is a challenging topic in computer vision due to its complexity and its important role in many applications such as human-computer and social interaction. This paper presents a Facial Expression Recognition (FER) method based on an automatic and more efficient facial decomposition into regions of interest (ROI). First, seven ROIs, representing more precisely facial components involved in expression of emotions (left eyebrow, right eyebrow, left eye, right eye, between eyebrows, nose and mouth), are extracted using the positions of some landmarks detected by IntraFace (IF).

International Review on Computers and Software (IRECOS), 2015

In this article, we describe a multi agent approach for supporting learning activities in a Virtu... more In this article, we describe a multi agent approach for supporting learning activities in a Virtual Learning Environment context. In order to assist teachers who monitor learning processes, viewed as a specific type of collaboration, the proposed system estimates a behavioral (sociological) profile for each student. This estimation is based on automatic analysis of students’ textual asynchronous conversations. The determined profiles are proposed to the teacher and may provide assistance to teacher during tutoring tasks. This paper specifically proposes definitions of the used sociological profiles, and presents the architecture of the Multi-Agent System (MAS) determining the profiles. This MAS relies on a fuzzy logic based strategy for representing the knowledge of the teacher. The system was experimented with students of the master “software quality” of the Ibn Tofail University. The results obtained show that the proposed approach is effective and gives satisfactory results.

2014 9th International Conference on Intelligent Systems: Theories and Applications (SITA-14), 2014

ABSTRACT In this article, we describe a new approach for the support and monitoring of an online ... more ABSTRACT In this article, we describe a new approach for the support and monitoring of an online collaborative learning process. The developed system features determination of behavioral sociological profiles of learners from the automatic analysis of textual asynchronous communications. This paper defines first the sociological profiles used in this model then describes the used method to analyses full text exchanged messages. We also present the indicators used for the determination of the learner's profiles. Then we present results of experimentations conducted on the developed system with students of "software quality" master of Ibn Tofail University, Kenitra, Morocco.

Proceedings of the New Challenges in Data Sciences: Acts of the Second Conference of the Moroccan Classification Society

In this paper, we propose an online course recommendation system applicable to a distance-learnin... more In this paper, we propose an online course recommendation system applicable to a distance-learning platform. (LMS) which assists and supports learners in knowledge tests. Most commercial LMS or open source as well as open source software do not include recommendation tools to guide learners during their learning process. Current online training devices, such as MOOCs, enlist thousands of learners, so tutors often feel powerless to build a synthetic representation of learner activity, to send recommendations, remarks to each learner and intervene in appropriate time. In an attempt to meet these needs, we propose an approach that aims to automatically analyze the learners' responses, which determine their level of knowledge. Basis on their answers, our system automatically offers personalized recommendations for the learners. We will then present the results of the implementation of the system developed as part of an experiment we conducted with 10,000 students enrolled on our distance-learning platform. The paper presents the results obtained and the perspectives for more research.