Search

Learning and teaching systems have seen fast transformations being increasingly applied in emerging formal and informal education contexts. Indeed, the shift to open learning environments is remarkable, where the number of students is extremely high. To allow a huge amount of learners gaining new knowledge and skills in an open education framework, the recourse to e-assessment systems able to cover this strong demand and respective challenges is inevitable. Facing Office skills as those most frequently needed in education and business settings, in this paper, we address the design of a novel assessment system for automated assessment of Office skills in an authentic context. This approach exploits the powerful potential of the Extensible Markup Language (XML) format and related technologies by transforming the model of both students’ documents and answers to an XML format and extracting from the teacher’s correct document the required skills as patterns. To assign a mark, we measure similarities between the patterns of the students’ and the teacher’s documents. We conducted an experimental study to validate our approach for Word processing skills assessment and developed a system that was evaluated in a real exam scenario. The results demonstrated the accuracy and suitability of this research direction.

The success of machine learning for automatic speech processing has raised the need for large scale datasets. However, collecting such data is often a challenging task as it implies significant investment involving time and money cost. In this paper, we devise a recipe for building largescale Speech Corpora by harnessing Web resources namely YouTube, other Social Media, Online Radio and TV. We illustrate our methodology by building KALAM’DZ, An Arabic Spoken corpus dedicated to Algerian dialectal varieties. The preliminary version of our dataset covers all major Algerian dialects. In addition, we make sure that this material takes into account numerous aspects that foster its richness. In fact, we have targeted various speech topics. Some automatic and manual annotations are provided. They gather useful information related to the speakers and sub-dialect information at the utterance level. Our corpus encompasses the 8 major Algerian Arabic sub-dialects with 4881 speakers and more than 104.4 hours segmented in utterances of at least 6 s.

The Label Propagation Algorithm (LPA) is a fastalgorithm for community detection. This algorithm has proved itsefficiency and scalability even in large social networks. It is basedonly on updating the label of each node by the most frequentlabel within its neighbors. However, in case of many maximallabels, a random choice is made. This reduces the performanceof the algorithm. In this paper, we propose a new version ofLPA, called Link Strength-based LPA (LS-LPA), that addressesthis issue. In our approach, we rank labels according to the linkstrength of each neighbor. The strength of links is quantified viak-neighborhood which includes common neighbors and multi-stepneighbors. Experiments on different networks with different sizesshow that our solution improves LPA's performance significantly.

The need for IT skills assessment and certification is indisputable for both persons and institutions. Thus, the recourse to automatic assessment systems able to answer the strong demand of such assessment or certification is inevitable. In this paper, we propose a new approach to automatically assess IT skills in authentic context. This approach relies on the potential of XML format and its related efficient technologies which allow the extraction of information from XML documents. We divide our assessment approach into two steps. In the first, we transform both Students documents and answers model to an intermediate representation using XML format for which some preprocessing is done. In the second step, we extract from the teacher's correct document the required skills as paths of the tree representation of the XML document. In order to assign a mark we measure similarities between these paths and those of the student document. We have conducted an experimental study to validate our approach. Thus, we have developped a tool based on the proposed approach to automatically assess Word Processing skills using Microsoft Office Word program. The evaluation of our tool is performed on a real exam with one hundred students. The results show the accuracy and the suitability of this solution for assessing IT skills.