PROJECTS
Current project is on Security and Privacy for Cloud Users
Full description coming soon.
Full description coming soon.
Yuruware Monitor
Yuruware Monitor is an application monitoring framework capable of monitoring application level metrics and detailed logs from a variety of applications including several CMS, CRM as well as linux servers, web servers and database servers. It has been developed as part of a suite of tools forming Yuruware, a successful startup company resulting from the spinoff of a research project conducted at NICTA (National ICT Australia). My involvement in this startup company helped me gain an invaluable entrepreneurial experience and hands-on experience with cloud technologies, business applications, as well as technical marketing skills. Yuruware was acquired by UniTrends, a large US company, for an undisclosed 8 digits figure (>$10 million) in May 2014 (read the news article from The Australian).
Yuruware Monitor is an application monitoring framework capable of monitoring application level metrics and detailed logs from a variety of applications including several CMS, CRM as well as linux servers, web servers and database servers. It has been developed as part of a suite of tools forming Yuruware, a successful startup company resulting from the spinoff of a research project conducted at NICTA (National ICT Australia). My involvement in this startup company helped me gain an invaluable entrepreneurial experience and hands-on experience with cloud technologies, business applications, as well as technical marketing skills. Yuruware was acquired by UniTrends, a large US company, for an undisclosed 8 digits figure (>$10 million) in May 2014 (read the news article from The Australian).
The ADAGE platform
The ADAGE platform (Ad-hoc DAta Grids Environments) is a user driven service oriented platform for efficiently gather, store, retrieve and process ad-hoc data grids from both a manager and an end-user perspective. Such platform can benefit various applications including in e-science and e-research.
I worked on both backend and frontend components of this project during the period 2007-2008. News and financial data has been the main application of ADAGE in collaboration with SIRCA, a leading provider of financial data and eResearch servicesI worked on both backend and frontend components of this project during the period 2007-2008. News and financial data has been the main application of ADAGE in collaboration with SIRCA, provider of financial data and eResearch services. The tool is now used as part of a CRC project (SmartServices).
Technical details: Two versions have been developed.
The ADAGE platform (Ad-hoc DAta Grids Environments) is a user driven service oriented platform for efficiently gather, store, retrieve and process ad-hoc data grids from both a manager and an end-user perspective. Such platform can benefit various applications including in e-science and e-research.
I worked on both backend and frontend components of this project during the period 2007-2008. News and financial data has been the main application of ADAGE in collaboration with SIRCA, a leading provider of financial data and eResearch servicesI worked on both backend and frontend components of this project during the period 2007-2008. News and financial data has been the main application of ADAGE in collaboration with SIRCA, provider of financial data and eResearch services. The tool is now used as part of a CRC project (SmartServices).
Technical details: Two versions have been developed.
- Version 1: Developed in HTML and Javascript with a backend in C# running on the mono framework as well as IIS.
- Version 2: Developed using GWT, SmartGWT, with a backend in JAVA and some C# services.
- Guabtni, A., Hundish, D., and Rabhi, F. A. 2010. A User-Driven SOA for Financial Market Data Analysis, accepted to be published in the international Journal of Enterprise Modelling and Information Systems Architectures
- Rabhi, F. A., Guabtni, A., and Yao, L. 2009. A data model for processing financial market and news data, International Journal of Electronic Finance, vol. 3, number 4, Inderscience Publishers, Geneva, Switzerland, (Oct. 2009), 387-403. http://dx.doi.org/10.1504/IJEF.2009.028978
- Rabhi, F. A., Rana, O. F., Guabtni, A., and Benatallah, B. 2008. A User-Driven Environment for Financial Market Data Analysis, FINANCECOM’08, 4th International Workshop on Enterprise Applications and Services in the Finance Industry, December 13th, Paris, France. http://dx.doi.org/10.1007/978-3-642-01197-9_5
A Cloud Based Distributed Query Processing for News Databases
This tool implements a novel approach balancing the workload of distributed range query processing on large replicated historical news databases using a cloud-based infrastructure. Processing range queries is a common operation in databases but can be computationally expensive when dealing with large data sets. Major database management systems allow to distribute the processing of a range query by splitting it across multiple replicated databases.
However, they assume a uniform distribution of data to achieve good performance. This assumption is very often not verified in real data, and more specifically in historical news data sets. Indeed, we demonstrate through a series of experiments on a large news database that classic approaches for splitting range queries do not optimally balance the workload across all database replicas. We propose an approach that takes into account the data density distribution when splitting a range query. This software is primary a cloud based testing environment for research experiments. It has been implemented using using PHP, HTML, CSS, using the Amazon EC2 API library in PHP. Performance stress testing using JMeter.
This tool implements a novel approach balancing the workload of distributed range query processing on large replicated historical news databases using a cloud-based infrastructure. Processing range queries is a common operation in databases but can be computationally expensive when dealing with large data sets. Major database management systems allow to distribute the processing of a range query by splitting it across multiple replicated databases.
However, they assume a uniform distribution of data to achieve good performance. This assumption is very often not verified in real data, and more specifically in historical news data sets. Indeed, we demonstrate through a series of experiments on a large news database that classic approaches for splitting range queries do not optimally balance the workload across all database replicas. We propose an approach that takes into account the data density distribution when splitting a range query. This software is primary a cloud based testing environment for research experiments. It has been implemented using using PHP, HTML, CSS, using the Amazon EC2 API library in PHP. Performance stress testing using JMeter.
A Graph Based Approach for Service Messages Correlation
This tool allows to build correlations of service interaction messages and present them in a graph based representation allowing a rapid identification of relevant correlation conditions. The tool takes as input a log file containing service interaction messages. The output of the tool is a hyper graph representing all the ways messages are correlated. Such graph is then used by the end user to quickly identify relevant correlation conditions.
I have been in charge of the research and development processes and a prototype version has been implemented and used as a proof of concept in a research paper under reviewing process.
Technical details: Developed using C# on the mono framework, HTML, CSS, ActionScript for flash-based graph visualization.
A research paper has been published in the proceedings of the international Conference on Advanced Information Systems Engineering CAiSE2011 held in London, United Kingdom, 20-24 June 2011.Title: "Using Graph Aggregation for Service Interaction Message Correlation".
Authors: Adnene Guabtni, Hamid R. Motahari-Nezhad, Boualem Benatallah.
This tool allows to build correlations of service interaction messages and present them in a graph based representation allowing a rapid identification of relevant correlation conditions. The tool takes as input a log file containing service interaction messages. The output of the tool is a hyper graph representing all the ways messages are correlated. Such graph is then used by the end user to quickly identify relevant correlation conditions.
I have been in charge of the research and development processes and a prototype version has been implemented and used as a proof of concept in a research paper under reviewing process.
Technical details: Developed using C# on the mono framework, HTML, CSS, ActionScript for flash-based graph visualization.
A research paper has been published in the proceedings of the international Conference on Advanced Information Systems Engineering CAiSE2011 held in London, United Kingdom, 20-24 June 2011.Title: "Using Graph Aggregation for Service Interaction Message Correlation".
Authors: Adnene Guabtni, Hamid R. Motahari-Nezhad, Boualem Benatallah.
Mind mapping as a service
This tool has been developed in collaboration with Stuart Clark, former student at CSE, UNSW. It consists of a tool providing mind-mapping as a service to be used by any web application. The possibilities of using this tool are endless. Indeed, one can represent any type of entity with properties and operations associated to it as well as entity relationships. Users can browse entities and perform actions on them using a mind-mapping style interface. An XML based API has been developed to allow third party applications to use the mind-mapping tool as a service, similarly to applications using Google Maps as a service.
A prototype version has been released as Open Source project on SourceForge and used as a proof of concept.
Project homepage - Screenshots - Try it online! - Demonstration video
Technical details: Developed using GWT, SmartGWT, JAVA, CSS.
A demo paper has been published and presented at the international World Wide Web conference WWW'2011 held in Hyderabad, India, 28th March - 1st April 2011.Title: "Embedding MindMap As a Service for User-Driven Composition of Web Applications"Authors: Adnene Guabtni, Stuart Clarke, Boualem Benatallah.
This tool has been developed in collaboration with Stuart Clark, former student at CSE, UNSW. It consists of a tool providing mind-mapping as a service to be used by any web application. The possibilities of using this tool are endless. Indeed, one can represent any type of entity with properties and operations associated to it as well as entity relationships. Users can browse entities and perform actions on them using a mind-mapping style interface. An XML based API has been developed to allow third party applications to use the mind-mapping tool as a service, similarly to applications using Google Maps as a service.
A prototype version has been released as Open Source project on SourceForge and used as a proof of concept.
Project homepage - Screenshots - Try it online! - Demonstration video
Technical details: Developed using GWT, SmartGWT, JAVA, CSS.
A demo paper has been published and presented at the international World Wide Web conference WWW'2011 held in Hyderabad, India, 28th March - 1st April 2011.Title: "Embedding MindMap As a Service for User-Driven Composition of Web Applications"Authors: Adnene Guabtni, Stuart Clarke, Boualem Benatallah.