Dr Francois Siewe

Job: Reader in Computer Science

Faculty: Technology

School/department: School of Computer Science and Informatics

Research group(s): Cyber Technology Institute (CTI) (Software Technology Research Laboratory (STRL))

Address: De Montfort University, The Gateway, Leicester, LE1 9BH, United Kingdom

T: +44 (0)116 257 7938

E: FSiewe@dmu.ac.uk

W: www.cse.dmu.ac.uk/~fsiewe/

 

Personal profile

I am a Reader in Computer Science in the Software Technology Research Laboratory (STRL) of the Faculty of Technology at De Montfort University (DMU), in Leicester in the UK. Before joining STRL, I was a research fellow on the EPSRC-funded project MELANGE on modelling the structure dependent colour properties of melange yarns in the Textile Engineering And Material (TEAM) research group at DMU. This followed my tenure as lecturer and vistiting researcher in the Institute of Technology of Lens at University of Artois in Lens, France. Prior to this, I was a fellow at the United Nation University/ International Institute for Software Technology (UNU/IIST) in Macau, where I worked on the Design Techniques for Reat-time systems (DeTfoRs) project. I was also a lecturer in the Department of Mathematics and Computer Science at the University of Dschang, Dschang, Cameroon.

Research group affiliations

Software Technology Research Laboratory (STRL)

Publications and outputs 

  • Agent-based Negotiation Approach for Feature Interactions in Smart Home Systems using Calculus of the Context-aware Ambient
    Agent-based Negotiation Approach for Feature Interactions in Smart Home Systems using Calculus of the Context-aware Ambient Alfakeeh, Ahmed S.; Al-Bayatti, Ali H.; Siewe, Francois; Baker, Thar Smart Home Systems (SHSs) provide several services which are tailored to different residents’ preferences. As a result, SHSs are highly exposed to undesirable interactions, known as feature interactions (FIs). FIs might occur as a result of a conflict in services’ goals or a conflict with residents’ preferences. Previous studies have proposed solutions based on applying priorities, in which some services or preferable features are disabled in favour of other services. Alternatively, the agent-based negotiation approach (ABNA) utilises agents and applies negotiation, enabling services with contrary features to work simultaneously. ABNA avoids applying priority between services or house residents’ preferences whenever a space for a compromise exists. The mechanism of ABNA is based on the use of a hierarchy of features based on their contribution to the function of the service or on the importance of these features to house residents. To achieve a compromise between conflicting services, ABNA models services and residents by using agents, and implements a negotiation algorithm that allows services with conflicting features to work simultaneously. This paper presents a description of ABNA with a formal specification of ABNA in the Calculus of Context-aware Ambient (CCA). This enables the formal analysis of ABNA by using the execution environment of CCA. The file attached to this record is the author's final peer reviewed version.
  • A novel algorithm for dynamic student profile adaptation based on learning styles
    A novel algorithm for dynamic student profile adaptation based on learning styles Nafea, Shaimaa M.; Siewe, Francois; He, Ying E-learning recommendation systems are used to enhance student performance and knowledge by providing tailor- made services based on the students’ preferences and learning styles, which are typically stored in student profiles. For such systems to remain effective, the profiles need to be able to adapt and reflect the students’ changing behaviour. In this paper, we introduce new algorithms that are designed to track student learning behaviour patterns, capture their learning styles, and maintain dynamic student profiles within a recommendation system (RS). This paper also proposes a new method to extract features that characterise student behaviour to identify students’ learning styles with respect to the Felder-Silverman learning style model (FSLSM). In order to test the efficiency of the proposed algorithm, we present a series of experiments that use a dataset of real students to demonstrate how our proposed algorithm can effectively model a dynamic student profile and adapt to different student learning behaviour. The results revealed that the students could effectively increase their learning efficiency and quality for the courses when the learning styles are identified, and proper recommendations are made by using our method. The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • On Recommendation of Learning Objects using Felder-Silverman Learning Style Model
    On Recommendation of Learning Objects using Felder-Silverman Learning Style Model Nafea, Shaimaa M.; Siewe, Francois; He, Ying The e-learning recommender system in learning institutions is increasingly becoming the preferred mode of delivery, as it enables learning anytime, anywhere. However, delivering personalised course learning objects based on learner preferences is still a challenge. Current mainstream recommendation algorithms, such as the Collaborative Filtering (CF) and Content-Based Filtering (CBF), deal with only two types of entities, namely users and items with their ratings. However, these methods do not pay attention to student preferences, such as learning styles, which are especially important for the accuracy of course learning objects prediction or recommendation. Moreover, several recommendation techniques experience cold-start and rating sparsity problems. To address the challenge of improving the quality of recommender systems, in this paper a novel recommender algorithm for machine learning is proposed, which combines students actual rating with their learning styles to recommend Top-N course learning objects (LOs). Various recommendation techniques are considered in an experimental study investigating the best technique to use in predicting student ratings for e-learning recommender systems. We use the Felder-Silverman Learning Styles Model (FSLSM) to represent both the student learning styles and the learning object profiles. The predicted rating has been compared with the actual student rating. This approach has been experimented on 80 students for an online course created in the MOODLE Learning Management System, while the evaluation of the experiments has been performed with the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results of the experiment verify that the proposed approach provides a higher prediction rating and significantly increases the accuracy of the recommendation. The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • A Novel Algorithm for Course Learning Object Recommendation Based on Student Learning Styles
    A Novel Algorithm for Course Learning Object Recommendation Based on Student Learning Styles Siewe, Francois; He, Ying; Nafea, Shaimaa Explosive growth of e-learning in the recent years has faced difficulty of locating appropriate learning resources to match the students learning styles. Recommender system is a promising technology in e-learning environments to present personalised offers and convey appropriate learning objects that match student inclinations. This paper, proposes a novel and effective recommender algorithm that recommends personalised learning objects based on the student learning styles. Various similarity metrics are considered in an experimental study to investigate the best similarity metrics to use in a recommender system for learning objects. The approach is based on the Felder and Silverman learning style model which is used to represent both the student learning styles and the learning object profiles. It was found that the K-means clustering algorithm, the cosine similarity metrics and the Pearson correlation coefficient are effective tools for implementing learning object recommender systems. The accuracy of the recommendations are measured using traditional evaluation metrics, namely the Mean Absolute Error and the Root Mean Squared Error.
  • ULEARN: Personalized Course Learning Objects Based on Hybrid Recommendation Approach
    ULEARN: Personalized Course Learning Objects Based on Hybrid Recommendation Approach Nafea, Shaimaa; Siewe, Francois; He, Ying The success of e-learning systems depends on their capability to automatically retrieve and recommend relevant learning content according to the preferences of specific learner profiles. Generally, e-learning systems do not cater for individual learners’ needs based on their profile. They also make it very difficult for learners to choose suitable resources for their learning. Matching the teaching strategy with the most appropriate learning object based on learning styles is presented in this paper, with the aim of improving learners’ academic levels. This work focuses on the design of a personalized e-learning environment based on a hybrid recommender system, collaborative filtering and item content filtering. It also describes the architecture of the ULEARN system. The ULEARN uses a recommender adaptive teaching strategy by choosing and sequencing learning objects that fit with the learners’ learning styles. The proposed system can be used to rearrange learning object priority to match the student’s adaptive profile and to adapt teaching strategy, in order to improve the quality of learning. The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • Formal Specification of a Context-aware Whiteboard System in CCA
    Formal Specification of a Context-aware Whiteboard System in CCA Atbaiga, N.; Siewe, Francois A context-aware whiteboard system provides a number of services in a smart classroom including registering students as they enter the classroom; logging students and lecturers in to the blackboard virtual learning environment at the beginning of each lecture and logging them out at the end of the lecture. This system also notifies students of their absence to a lecture and maintains a list of attendance automatically. Using information from the timetable, it is aware of the lectures that are scheduled to take place in the classroom and the students that are allowed to attend these lectures. Finally, it allows students and lecturers to interact with teaching materials such as lecture slides and videos stored in the blackboard virtual learning environment. This paper proposes a formal specification of the white board system in the Calculus of Context-aware Ambients (CCA in short). This enables the formal analysis of the white board system using the execution environment of CCA. Some important properties of a classroom white board system have been validated as a proof of concept.
  • ULEARN: Personalized course Learning Objects based on Hybrid Recommendation approach
    ULEARN: Personalized course Learning Objects based on Hybrid Recommendation approach Nafea, Shaimaa; Siewe, Francois; He, Ying Adaptive e-learning recommender system is observed one of the exciting research discipline in the education and teaching throughout the past few decades, since, the learning style is specific for each student In reality from the knowledge of his/ her learning style; matching teaching strategy with the most appropriate learning object is present to better return on learner academic level. This work focuses on the design of a personalized e-learning environment based on hybrid recommender system based on collaborative filtering and item content filtering as well as architecture of ULEARN system. ULEARN recommended adaptive teaching strategy by choosing and sequencing learning objects fitting with the learners’ learning styles. The proposed system can be used to rearrange learning object priority that matches student adaptive profile and teaching strategy in order to improve the quality of learning.
  • ULEARN: Personalised Learner’s Profile Based On Dynamic Learning Style Questionnaire
    ULEARN: Personalised Learner’s Profile Based On Dynamic Learning Style Questionnaire Nafea, Shaimaa; Siewe, Francois; He, Ying E-Learning recommender system effectiveness re- lies upon their ability to recommend appropriate learning con- tents according to the learner learning style and preferences. An effective approach to handle the learner preferences is to build an efficient learner profile in order to gain adaptation and individualisation of the learning environment. It is usually necessary to know learning style and preferences of the learner on a domain before adapting the learning process and course content. This study focuses on identifying the learning styles of students in order to adapt the learning process and course content. ULEARN is an adaptive recommender learning system designed to provide learners with personalised learning environment such as course learning objects that match their adaptive profile. This paper presents the algorithm used in ULEARN to reduce dynamically the number of questions in Felder-Silverman learning style ques- tionnaire used to initialise the adaptive learner profile. Firstly, the questionnaire is restructured into four groups, one for each learning style dimension; and a study is carried out to determine the order in which questions will be asked in each dimension. Then an algorithm is built upon this ranking of questions to calculate dynamically the initial learning style of the user as they go through the questionnaire. The file attached to this record is the author's final peer reviewed version.
  • A Multi-Layer Framework for Quality of Context in Ubiquitous Context-Aware Systems
    A Multi-Layer Framework for Quality of Context in Ubiquitous Context-Aware Systems Al-Shargabi, A. A.; Siewe, Francois This paper proposes a novel framework for Quality of Context (QoC) in context-aware systems. The main innovative features include: (1) a new definition that generalizes the notion of QoC; (2) a novel multilayer context model; (3) a novel model of QoC that introduces new quality parameters; (4) a novel mechanism to define QoC policy by assigning weights to QoC parameters using a multi-criteria decision-making technique; (5) and a novel quality control algorithm that handles context conflicts, context missing values, and context erroneous values. Our frameworkis implemented in MatLab and evaluated using a case study of a flood forecast system. The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.
  • An Adaptive Learning Ontological Framework Based on Learning Styles and Teaching Strategies
    An Adaptive Learning Ontological Framework Based on Learning Styles and Teaching Strategies Nafea, Shaimaa; Siewe, Francois; He, Ying Ontology are increasingly being used in a variety of applications, and particularly in adaptive e-learning. They have the potential to enable developers to create adaptive course content for specified domains. E-learning applications are thus able to use technology and educational content in order to generate content that matches the student's capabilities and knowledge. This personalises learning, rather than assuming that "one-size-fits-all" and providing all learners with the same content, which is what the majority of e-learning systems do. This study introduces a new approach that takes into account the fact that each learner has an individual learning style and needs. The approach enables to adapt the course content, teaching strategy and learning objects so that they correspond to each student’s learning styles. This is achieved with the use of artificial intelligent in the form of an ontology and rule-based reasoning. The proposed system takes some of the key design aspects such as extensibility, reusability, and maintainability into consideration in order to enhance performance of adaptive course content recommendation.

Click here to view a full listing of Francois Siewe's publications and outputs.

Research interests/expertise

Computer Security

Pervasive Systems

Formal Methods

Process Calculi

Formal Verification

Qualifications

PhD in Computer Science, De Montfort University, Leicester, UK, 2005

Doctorat de Troisième Cycle, Université de Yaoundé I, Yaoundé, Cameroon, 1997

Msc in Computer Science, Université de Yaoundé I, Yaoundé, Cameroon, 1992

Bsc in Maths & Computer Science, Université de Yaoundé, Yaoundé, Cameroon, 1990

Courses taught

Pervasive Systems

Honours and awards

Research Journal of Textile and Apparel (RJTA) Excellent Paper Award 2011

Current research students

First supervisor for:

Ahmed Mohammed Alalshuhai
Abdulgader Zaid Almutairi
Saud Maqed Almutairi
Muslit Awadh Alotaibi
Asma Abdulghani Qasem Al-Shargabi
Elena Chernikova
Abdulghani Mahmoud Suwan

Second supervisor for:

Abdullah Shawan Alotaibi
Amr Mohsen Jadi

Francios-Siewe

Search Who's Who

 
News target area image
News

DMU is a dynamic university, read about what we have been up to in our latest news section.

Events target area image
Events

At DMU there is always something to do or see, check out our events for yourself.

Mission and vision target area image
Mission and vision

Read about our mission and vision and how these create a supportive and exciting learning environment.