E-learning environments are part of the self-learning ecosystem of which self-regulation is one of the main pillars. One of the key factors contributing to dropping out in an e-learning environment is the lack of self-regulatory skills. This skill is often absent in many learners, leading them to abandon their training. In this paper we propose a methodology, chatbot architectural design, allowing to help the learner to self-regulate in his/her learning by supporting him/her through a chatbot within the Moodle platform, which constitutes a metacognitive virtual assistant, to provide metacognitive messages. The aim is to improve the quality of learning in the e-learning environment.

In recent years, the development of recommendation systems has aroused growing interest in many areas, especially in e-learning. However, the lack of support and personalization in this context leads the learners to lose their motivation to continue in the learning process. To overcome this problem, we focus on adapting learning resources to the needs of learners using a recommender system. Thus, most of the existing studies in this area don’t take into account the differences in the characteristics of learners. This problem can be mitigated by incorporating additional learner information into the recommendation process. Additionally, many recommender techniques experience cold start and rating sparsity issues. This article presents our proposed recommendation system in order to provide learners with appropriate learning resources to follow the learning process and maintain their motivation. We propose hybrid recommender system that combines recommendation techniques to solve the problem of retrieving relevant learning resources for learners and that can alleviate both the cold-start and data sparsity problems. Using this hybridization between different techniques is useful in the personalization of the learner’s profile. We consider different learner characteristics such as their preferences, their goal, and their level of knowledge, hence resulting in the generation of more accurate recommendations.

Since the beginning of the COVID-19 pandemic, many countries have adopted online education as an alternative to face-to-face courses. This has increased awareness of the importance of analyzing learning data left by students to improve and evaluate the learning process. This article presents a new tool, named TaBAT, created to work with different LMSs in the form of dashboards accessible online and allowing teachers to monitor the progress of their learners and at the same time and allow teachers to track the progress of their learners, while allowing learners to develop self-regulation skills and visualize their learning process. The results of a study conducted show that TaBAT helped learners to increase their progress and spend more time in the online course.

Since learning in higher education is increasingly taking place online, the multiplication of web-based educational content, learning management systems (LMSs) and collaborative communication platforms have generated a large volume of data on learners and their learning activities. In recent years, interest has been growing in analyzing this data to support real-time decision making and improve the learning experience. This paper presents the results of a study conducted in higher education in Morocco, which evaluates a learning analysis dashboards (LADs) for both teachers and learners. The study shows that the dashboard, called TABAT, allowed a synthetic visualization of learning progress in courses and led to improved student engagement and success rates.

E-learning constitutes an alternative to face to face traditional educational system. Unfortunately, many learners stop their online learning at its beginning. This paper aims system at exploring and identifying the critical success factors (CSFs) to predict learners' satisfaction, continuance and success in on line platform learning. We also proposed a hybrid model that classify success factors of online learning in four dimensions include the learner characteristics, the instructor characteristics, the system specificities and the social influence

We experimentally test the effectiveness and reliability of the accelerometer-based tilt sensor as an interaction device for pointing and scrolling tasks on handheld devices. The touch screen and stylus based interfaces on these devices requires using two hands: one for holding the device and the other for driving the stylus. In our experiment, the participants use only one hand to interact with the same interfaces. In addition to freeing one hand, using the tilt sensor doesn't hide a part of the device's tiny screen. Two different interfaces are used in the experiment: one for testing pointing and targeting without scrolling and the other is for testing them with scrolling. Using the experiment results we developed a model for the prediction of the execution time for both tasks.

We present a platform for mobile text entry evaluation. The platform comprises two parts: the experimentation part and the analysis and design support part. The experimentation part is available in two environments: it can run, on top of the .NET framework, on a mobile device or on the desktop computer. This part is designed to save time and effort in the empirical evaluation process and to enhance the comparison of results from different experiments. The analysis and design support part is a collection of tools designed to support the evaluation. Our goal is to help in carrying out evaluation and to make a step on the way to the standardization of the evaluation process.

In recent works, we have developed a text input method based on analogy with capital Latin characters and on the decomposition of characters into basic shapes. It has been designed to be universal and to allow for typing text with only one keystroke per character. In this paper we present a new implementation of this method for users with motor impairments. It uses the same principles in addition to some common techniques used in the Augmentative and Alternative Communication (AAC) domain. The new solution has been tested with a person with Locked-In Syndrome and the results are promising.

In this paper, we describe the latest version of SIBYLLE, an AAC system that permits persons suffering from severe physical disabilities to enter text with any computer application and also to compose messages to be read out by a speech synthesis module. The system consists of a virtual keyboard comprising a set of keypads which allow entering characters or full words by a single-switch selection process. It also comprises a sophisticated word prediction component which dynamically calculates the most appropriate words for a given context. This component is auto-adaptive, i.e. it learns on every text the user has entered. It thus adapts its predictions to the user's language and the current topic of communication as well. So far the system works for French, German and English. Earlier versions of SIBYLLE have been used since 2001 in the Kerpape rehabilitation center (Brittany, France).

This paper discusses the issue of how a speech understanding system can be made robust against spontaneous speech phenomena (hesitations and repairs) as well as achieving a detailed analysis of spoken French. The Romus system is presented. It implements speech understanding in a two-stage process. The first stage achieves a finite-state shallow parsing that consists in segmenting the recognized sentence into basic units (spokenadapted chunks). The second one, a Link Grammar parser, looks for inter-chunks dependencies in order to build a rich representation of the semantic structure of the utterance. These dependencies are mainly investigated at a pragmatic level through the consideration of a task concept hierarchy. Discussion about the approach adopted, its benefits and limitations, is based on the results of the system's assessment carried out under different linguistic phenomena during an evaluation campaign held by the French CNRS.

In this paper, we present a novel hybrid keyword spotting system that combines supervised and semi-supervised competitive learning algorithms. The rst stage is a S-SOM (Semi-supervised SelfOrganizing Map) module which is speci cally designed for discrimination between keywords (KWs) and non-keywords (NKWs). The second stage is an FDVQ (Fuzzy Dynamic Vector Quantization) module which consists of discriminating between KWs detected by the rst stage processing. The experiment on Switchboard database has show an improvement of about 6% on the accuracy of the system comparing to our best keyword-spotter one.

We present a novel hybrid keyword spotting system that combines supervised and unsupervised competitive learning algorithms. The first stage is a SOFM (self-organizing feature maps) module which is specifically designed for discriminating between keywords (KWs) and non-keywords (NKWs). The second stage is a FDVQ (fuzzy dynamic vector quantization) module which consists of discriminating between KWs detected by the first stage processing. The results show an improvement of about 9% on the accuracy of the system comparing to our standard one.

In this paper, three techniques are introduced to improve the DVQ algorithm. The first one consists in an automatic algorithm to determine the threshold oe, a function of the minimum class variance, by a cross validation procedure. The second improvement consists in a new adaptive learning rule based on a spatial geometry considerations. The algorithm is proposed to reduce an instability phenomenon which may appear during learning. Finally a criterion of neurons annihilation is proposed to remove unnecessary elements from the system to ensure network stability. Some experiments on real speech data are presented to show the effects of these three techniques on the network properties, including the learning time duration, the number of references and the recognition rate in each case. Key Words : Automatic Speech Recognition, Artificial Neural Network, Incremental Learning, Neurons Annihilation. TADJ Chakib, EuroSpeech, 1993 1. INTRODUCTION In recent years, there has been considerabl...

Learning vector quantization (LVQ) is a nearest neighbor classifier very close to the self-organizing feature map classifier. A novel method called dynamic vector quantization (DVQ) is proposed for improving the ability to generalize and the learning and testing speed. DVQ is evaluated on speech data and synthetic data. DVQ always gives best results with fewer reference vectors than LVQ2. On speech experiments, DVQ shows an improvement of about 5% in the recognition rate, and the learning speed is three times faster.>