BBBoumediene Belkhouche

Boumediene Belkhouche

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Boumediene Belkhouche has a total of 55 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Definition and Application of a Culture-Based Evaluator Selection Model Integrating Profile Theory and Machine Learning

    by Khaled Hamdan, Boumediene Belkhouche

    Global cross-cultural evaluation highlights the disruptive impact of culture on software evaluation. We propose a culture-based model for selecting software evaluators, using profile theory and machine learning to ensure that evaluators are both technically competent and aligned with the cultural context. Using profile theory, we develop a formal model that captures characteristics required for contextually effective software evaluators. The defining characteristics include cultural, organizational, technical, and individual attributes, as well as their interrelationships. A profile is defined using two formulas that integrate the numerical values of these characteristics to derive capability and compatibility properties. Our approach includes identification of evaluator characteristics, evaluator attributes selection, data analysis, and formulation of a model based on profile theory. Our model can be utilized to construct profiles by evaluating profile components, comparing profile instances, and selecting relevant profiles. Machine learning methods, including Multilayer Perceptron, Random Forest, Naive Bayes, tree-based regression, Support Vector Machines, and Bayesian Networks, are applied to assess predictive performance and identify influential evaluator and cultural attributes to strengthen the empirical foundation of the model. Dimensionality reduction and clustering techniques reveal distinct evaluator profiles, while regression models demonstrate strong explanatory power for evaluator capability. A CASE tool automates the evaluator selection process and enables project managers to compare profiles. A graphical user interface provides a user-friendly platform for computing and visualizing capability and compatibility metrics. The results demonstrate that integrating cultural and cognitive attributes with machine learning analytics enhances evaluator selection, improves consistency, and supports culturally aligned software evaluation in global contexts.

  • Open Access

    Framework for Personalized Content Recommendations to Support Informal Learning in Massively Diverse Information Wikis

    by Heba M. Ismail, Boumediene Belkhouche, Saad Harous

    Even though, most of the early efforts in personalized learning focused on formal learning, there is a growing undeniable demand for personalized support for informal learning. Wikis among other information-oriented platforms are experiencing an increasing attention for informal learning, especially Wikipedia. Link-based navigation and keyword-based search methods used on wiki environments suffer from many limitations. To support informal learning on these environments, it is important to provide easy and fast access to relevant content. However, the massive diversity of unstructured content and user base on these environments pose major challenges when designing recommendation models. To the best of our knowledge, no effective personalized content recommendation approach has yet been defined to support informal learning on wikis. We propose an effective personalized content recommendation framework (PCRF) in addition to an evaluation framework that can be used to evaluate the impact of personalized content recommendations on informal learning from wikis. PCRF implements an efficient structural recommendation model by integrating fuzzy thesauri with adaptive users’ interest models generated using structural analysis of topical navigational graphs. We design user studies with multiple strategies and treatments to evaluate the effectiveness of the framework and assess the impact of recommendations on informal learning. Experiments show that PCRF generates highly relevant recommendations adaptive to changes in users’ interests using the HARD model with MAP@k scores 86.4–100%. An evaluation of informal learning revealed that users of Wikipedia with personalized support could achieve higher scores on a conceptual knowledge assessment with an average score of 14.9 compared with 10.0 for users who used the encyclopedia without recommendations. Results confirm that PCRF can effectively support informal learning on wikis and similar environments.


  • A Culture-Based Profile Model of Software Evaluators

    by Khaled Hamdan, Boumediene Belkhouche, Peter Smith

  • Semantic Twitter sentiment analysis based on a fuzzy thesaurus

    by Heba M. Ismail, Boumediene Belkhouche, Nazar Zaki


  • A Software Model Supporting Smart Learning

    by Shamsa Abdulla Al Mazrouei, Nafla Saeed Al Derei, Boumediene Belkhouche


  • The Influence of Leadership and Work Culture on Software Cost Effort

    by Boumediene Belkhouche, Khaled Hamdan, Peter Smith

    Globalization is adding more dimensions to software effort estimation process. The notions of leadership and culture carry with them highly variable assumptions, and thus, must be explicitly modeled. A new model that incorporates leadership and culture is proposed, elaborated and validated. A survey was undertaken to determine the impact of culture and its effect on the software development process in the areas of project team timeliness, collaboration and team work, leadership characteristics, cultural intelligence, motivation and communication.. The use of the Bootstrap method for estimating the effort involved in a given project, along with analogies using real historical data, demonstrates the effectiveness of this approach in surmounting difficulties in describing abstract quantitative variables. Our approach is tested on a cluster sample dataset of 41 cases (projects) collected in 2007 from more than 20 organizations. The results show that the inclusion of leadership and culture in the cost estimation model improves the accuracy of software cost estimation

  • Image abstraction for improved semantic retrieval accuracy and reduced space-time complexities

    by Anas Y. Boubas, Saad Harous, Boumediene Belkhouche

    There have been many developments in content-based image retrieval (CBIR) recently. While retrieval quality of such systems has improved dramatically in the last decade, the complexities involved made most available systems impractical in real-life use. These complexities are a combination of space use and computational time required. Most approaches available to reduce these two complexities are a trade-off with retrieval accuracy and quality. In this work, we demonstrate a novel abstraction system based on Singular Value Decomposition (SVD), which is able to reduce the complexities of image retrieval systems, while at the same time enhancing the retrieval quality. We used an implementation of a very recent content-based image retrieval system named IPSILON [11] to demonstrate how quality and complexity are greatly enhanced.


  • What is a modern Arabic dictionary?

    by B. Belkhouche, H. Harmain, H. Al Taha, L. Al Najjar

    A major concern our research addresses is how to enhance the children's use of the Arabic dictionary to encourage them to independently learn more vocabulary. Even though the first Arabic dictionary was developed almost 1500 years ago, the state of the art in Arabic dictionaries has not made much progress and falls far behind dictionaries for other languages. We propose a framework for building dictionaries that are game-like in their level of interactivity, and thus provide an immersive environment for children to enthusiastically explore more concepts, images, and vocabulary.


  • GENESTEM: A novel approach for an Arabic stemmer using genetic algorithms

    by Anas Boubas, Leena Lulu, Boumediene Belkhouche, Saad Harous

    We use genetic algorithms and pattern matching to generate a morphological analyzer for Arabic verbs. Our approach consisted of developing general verb patterns and then applying these patterns to derive morphological rules. Except for some rare ambiguous cases, the resulting morphological analyzer is capable of recognizing all instances of verbs.


  • Formal specification and simulation of the robot path planner

    by M. Yassine Belkhouche, Boumediene Belkhouche

    Simulation software plays a very important role in education and research applications. They provide a safe, easy and flexible environment for developing and testing new methods. However developing a simulation software that reflects the exact behavior of the real system is a difficult and complicated task. In this paper we used hybrid process algebra to build a formal model that describes the exact behavior of the path planner module of autonomous mobile robots. The robot path planner is described by a powerful recursive process. The robot moves in four modes heading-regulation, move-to-goal, move-to-inter-goal, and obstacle-avoidance mode. Each motion mode is described by a process. The overall behavior of the path planner is the result of the communication between the four processes. The developed model is then implemented as simulator for mobile robots.