Sarra Bouhenni

I work on designing large scale parallel and distributed machine learning algorithms for streaming data.

Ubiquity.AI helps Ad Networks to increase their ROI by removing ad decisioning bottlenecks thanks to a proprietary Decision as a Service platform that increases their margin up to 300% while preserving the scale. Unlike solutions that rely on constant manual tuning or general-purpose Machine Learning not fitted for all KPIs, Ubiquity.AI empowers Ad Networks to grow their margin and customer base without growing their operational costs.

I worked on distributed algorithms of big graphs and big data processing platforms.

My research interests include: Distributed algorithms, Graph algorithms, Graph Pattern Matching, ML Applied to graphs, graph data models.

I delivered the programming course Big Data - Scala in the post-graduation training PGS. It covers: basic syntax of Scala programming language, functional programming with Scala, and Scala collections.

I also delivered the introductory course Frameworks for big graphs - GraphX.

I worked on improving the logging system of the VTC application Yassir. My responsitibilies included mainly:

• Extraction of the log events available at a Kafka server using a logger implemented with NodeJs.
• Transformation of the events to compresenhive information to be stored on an Elasticsearch server.
• Creation of dashboards for the visualization and reporting of logs using Kibana.

In my final year’s engineering project, I worked on the design and development of a monitoring system for predicting plant diseases in Algeria, using Incremental Machine Learning algorithms. My responsibilities in this project included:

• Collecting data on plant disease prediction for the two diseases Fusarium Head Blight and Potato Mildiew.
• Design of a new algorithm for plant disease prediction based on incremental learning.
• Design and development of a backend for disease prediction based on current weather conditions, using Python and Django.
• Design and development of a web application, using Angular, for monitoring and visualization of the disease spread across Algerian crops.

I worked in my PhD on the problem of Graph Pattern Matching (GPM) in the context of large-scale distributed graphs. PhD Thesis dissertation can be downloaded via this link: Parallel and Distributed Algorithms for Pattern Matching in Big Graphs

My final year’s project was about the development of a plant disease forecasting platform. Thesis dissertation can be downloaded via this link: APDM : vers une plateforme intelligente pour la prevision des maladies végétales