Biography

Hi! I’m Chemseddine Berbague and welcome to my little corner of the internet. I am a developer and a data scientist. I have got my Ph.D. from the University of Badji Mokhtar-Annaba, Algeria, in 2021. Under my research, I investigate the development of new ideas and solutions to improve the performances of recommender systems, which are critical in any customization system. I am an author of several international conferences and journals, in which I treated the multi-objective recommendation quality problem by developing better collaborative-based filtering techniques. My current interest is improving the evolutionary algorithms using deep learning and solving more pressing real recommendation problems such as the issues related to health, education, and news applications.

Education

PhD in Computer Science, 2021.

Durnig my doctoral research, we interested by improving the recommendation quality in e-commerce applications. In effect, we proposed a multi-objective genetic-based clustering algorthim to balance diversity, novelty, and relevancy of recommendation, which are conflicting quality metrics. In more details, we used performant optmization and data reduction techniques to group together the users into similar and diverse clusters. Our proposed solutions have achieved comparative results in comparison to similar appraoches. You find on Slideshare a more detailed presentation of our proposition or you can contact me to get a copy of my thesis.

Badji Mokhtar University, Annaba, Algeria

MSc in Science and Technologies of Information and Communication, 2015.

In my Master's project, we have developed a visual Java application to explore, query, filter, and explore semantic knowledge from OWL/RDFS ontologies using the querying language SparQL, and the association rules technique. The results, then, were validated and re-inserted to enrich the information system. You find on Slideshare a more detailed presentation of our project. Also, we have shared the code of this project on GitHub.
Badji Mokhtar University, Annaba, Algeria

Research Interest

Data reduction & optimization

evolutionary algorithms; and statistical-based data reduction techniques.

Recommendation systems

recommendation evaluation; rating prediction; recommednation ranking; re-ranking; and scalability.
Semantic web & open linked data

knowledge representation; vectorization; mining; classification; and exploration.

Experiences

Academic experiences

Teacher

During 3 years, I have worked as a part-time teacher and gave Matlab practical activities, in both science and techniques department (ST), and computer and mathematics department (MIAS) at the university Badji-Mokhtar, Annaba, Algeria.
Reviewer

RecSys 2020,ECAI 2020, TheWebConf 2020, TheWebConf 2019, and ACM UMAP 2019
Researcher

An active researcher within the laboratory LabGed, computer science department, Annaba, Algeria.

Professional experiences

Data Management Engineer

As an Odoo developer, and a business analyst, at Nechma Group Company, I have collaborated within a team to analyze the needs of the company and proposed adequate solutions for automatizing the information process. In the projects I worked on, I had a cumulative experience with Python, XML, business analysis, and team work.
Freelance Web Developer

I have worked as an independent web developer on some web applications. In such projects, I used known web technologies and frameworks such as react, VueJS, and bootstrap. Have a look on some of my open source samples.

Research work

Selected publication

An Overlapping Clustering Approach for Precision, Diversity and NoveltyAware Recommendations.

Chemseddine Berbague, Karabadji Nour El-Islem, Seridi Hassina, Panagiotis Symeionidis, Yannis Manopolous, Wajdi Dhifli.

Check My Google Scholar DOI

Development skills

Programming languages

Java Python Matlab C#

Libraries & tools

Surprise Panda Keras Mahout Weka RankSys