Jamal Bentahar | Concordia University (Canada) (original) (raw)

Uploads

Papers by Jamal Bentahar

Computers & Electrical Engineering, Oct 1, 2019

Lecture Notes in Computer Science, 2023

IEEE transactions on neural networks and learning systems, Jul 1, 2023

Probabilistic topic models are considered as an effective framework for text analysis that uncove... more Probabilistic topic models are considered as an effective framework for text analysis that uncovers the main topics in an unlabeled set of documents. However, the inferred topics by traditional topic models are often unclear and not easy to interpret because they do not account for semantic structures in language. Recently, a number of topic modeling approaches tend to leverage domain knowledge to enhance the quality of the learned topics, but they still assume a multinomial or Gaussian document likelihood in the Euclidean space, which often results in information loss and poor performance. In this article, we propose a Bayesian embedded spherical topic model (ESTM) that combines both knowledge graph and word embeddings in a non-Euclidean curved space, the hypersphere, for better topic interpretability and discriminative text representations. Extensive experimental results show that our proposed model successfully uncovers interpretable topics and learns high-quality text representations useful for common natural language processing (NLP) tasks across multiple benchmark datasets.

IGI Global eBooks, May 25, 2011

2022 26th International Conference on Pattern Recognition (ICPR), Aug 21, 2022

Advances in systems analysis, software engineering, and high performance computing book series, 2013

There is a growing interest in the research and industry communities to examine the possible weav... more There is a growing interest in the research and industry communities to examine the possible weaving of social elements into Web services-based applications. This interest is backed by the widespread adoption of Web 2.0 technologies and tools developed using various online means such as social networks and Blogs. Social Web services incorporate the result of this weaving and are concerned with establishing relationships with their peers like people do daily. This chapter reviews the recent developments in this new topic and identifies new research opportunities and directions that are still unexplored such as security, engineering, reputation, trust, and argumentation.

arXiv (Cornell University), Apr 26, 2021

Composite services of type SaaS run in dynamic distributed environments where the quality of serv... more Composite services of type SaaS run in dynamic distributed environments where the quality of service (QoS) properties of the constituent services may change during execution. To face such dynamism and volatility, adaptation of composite SaaS to the runtime changes is a vital requirement. Recent research focused on centralized environments which are impractical for dynamic composition that requires distributed settings. To address this challenge, this paper proposes a distributed approach for composite SaaS adaptation using feature selection through applying the master/slave pattern. Slaves locally monitor the distributed constituent services and send performance information to the master, which in its turn reconfigures the composite services to provide the expected QoS and monitors the overall performance. Since adapting a composite SaaS to be QoS-optimal depends on multiple criteria according to the selected features, e.g., performance and cost, we model the adaptation process as a multi-objective optimization problem and then propose a genetic algorithm to compute the Pareto-optimal set of solutions for this problem. Experimental results show that our approach is efficient in distributed and large-scale environments compared to the centralized approach.

Service providers and users are located at the opposite ends of the Gigabit passive optical acces... more Service providers and users are located at the opposite ends of the Gigabit passive optical access network (GPON) channel [1] and hence fair resource allocation and control for both service providers and users can be very challenging. This paper introduces an integrated scheme for optimizing resource allocation in multi-operator GPON access networks.

ACM Transactions on Internet Technology, Nov 9, 2019

Computers & Electrical Engineering, Oct 1, 2019

Lecture Notes in Computer Science, 2023

IEEE transactions on neural networks and learning systems, Jul 1, 2023

Probabilistic topic models are considered as an effective framework for text analysis that uncove... more Probabilistic topic models are considered as an effective framework for text analysis that uncovers the main topics in an unlabeled set of documents. However, the inferred topics by traditional topic models are often unclear and not easy to interpret because they do not account for semantic structures in language. Recently, a number of topic modeling approaches tend to leverage domain knowledge to enhance the quality of the learned topics, but they still assume a multinomial or Gaussian document likelihood in the Euclidean space, which often results in information loss and poor performance. In this article, we propose a Bayesian embedded spherical topic model (ESTM) that combines both knowledge graph and word embeddings in a non-Euclidean curved space, the hypersphere, for better topic interpretability and discriminative text representations. Extensive experimental results show that our proposed model successfully uncovers interpretable topics and learns high-quality text representations useful for common natural language processing (NLP) tasks across multiple benchmark datasets.

IGI Global eBooks, May 25, 2011

2022 26th International Conference on Pattern Recognition (ICPR), Aug 21, 2022

Advances in systems analysis, software engineering, and high performance computing book series, 2013

There is a growing interest in the research and industry communities to examine the possible weav... more There is a growing interest in the research and industry communities to examine the possible weaving of social elements into Web services-based applications. This interest is backed by the widespread adoption of Web 2.0 technologies and tools developed using various online means such as social networks and Blogs. Social Web services incorporate the result of this weaving and are concerned with establishing relationships with their peers like people do daily. This chapter reviews the recent developments in this new topic and identifies new research opportunities and directions that are still unexplored such as security, engineering, reputation, trust, and argumentation.

arXiv (Cornell University), Apr 26, 2021

Composite services of type SaaS run in dynamic distributed environments where the quality of serv... more Composite services of type SaaS run in dynamic distributed environments where the quality of service (QoS) properties of the constituent services may change during execution. To face such dynamism and volatility, adaptation of composite SaaS to the runtime changes is a vital requirement. Recent research focused on centralized environments which are impractical for dynamic composition that requires distributed settings. To address this challenge, this paper proposes a distributed approach for composite SaaS adaptation using feature selection through applying the master/slave pattern. Slaves locally monitor the distributed constituent services and send performance information to the master, which in its turn reconfigures the composite services to provide the expected QoS and monitors the overall performance. Since adapting a composite SaaS to be QoS-optimal depends on multiple criteria according to the selected features, e.g., performance and cost, we model the adaptation process as a multi-objective optimization problem and then propose a genetic algorithm to compute the Pareto-optimal set of solutions for this problem. Experimental results show that our approach is efficient in distributed and large-scale environments compared to the centralized approach.

Service providers and users are located at the opposite ends of the Gigabit passive optical acces... more Service providers and users are located at the opposite ends of the Gigabit passive optical access network (GPON) channel [1] and hence fair resource allocation and control for both service providers and users can be very challenging. This paper introduces an integrated scheme for optimizing resource allocation in multi-operator GPON access networks.

ACM Transactions on Internet Technology, Nov 9, 2019

Log In