Bio & Skills

My name is Mohamed Eldesouki Mohamed, and I am currently a research associate at the Qatar Computing Research Institute (QCRI) working with Dr. Kareem Darwish in the Arabic Language Technologies (ALT) group. I have research and development experience in the fields of Natural Language Processing (NLP) and Information Retrieval (IR) in which I earned a master’s degree. Also, I have several publications related to IR and NLP published in top-tier conferences.

Joining a prestigious research institute like QCRI contributed substantially in crafting my research personality. It gave me the opportunity to learn the craft of research through participating in developing state-of-the-art NLP & deep learning technologies and through publishing in top-tier conferences. Furthermore, I gained practical experience from several projects namely FARASA (a fast and accurate NLP toolkit for Arabic text), Dialectal Arabic Tools, ClassStrength, and Live Arabic Dialects Identification of Arabic Speech.

Previously as research assistant at Cairo University, I had the responsibility to prepare and teach two courses, namely introduction to programming using Python and fundamentals of NLP. Meanwhile, I participated in developing APC (Arabic Processing Cog) a python text processing package. I also developed WebCS (a news aggregator that crawl more than 120 Arabic news sites and automatically recognize the title, body, author(s) and publish date).

In addition to academia, I worked in industry as a senior R&D engineer for more than 6 years which has sharpened my skills in using a vast spectrum of programming languages and technologies to build real-world large-scale projects and applications leveraging machine and deep learning. Some of the models that I have used are Linear & Logistic regression, CRF, SVM, CNN, LSTM, and word embeddings. Some of these I have implemented myself and I used tools such as Keras and TensorFlow for others..

Professional Experience
Dec 2015 - Sep 2019
Research Associate

Qatar Computing Research Institute (QCRI)

Doha, Qatar

I am working on Arabic language technologies with a focus on dialectal Arabic Processing and speech-enabled dialect identification.

Jan 2015 - Dec 2015
Senior Research Engineer

OMS Company

Cairo, Egypt

Aug 2013 - Dec 2014
Senior Research Engineer

Taya IT Company

Cairo, Egypt

I Built the core technology behind TAPS (Taya Arabic Processing Suite), that includes morphological analysis, Named Entity Recognition and Autocomplete and spelling correction.

Jul 2009 - Jun 2012
Research Assistant

Cairo University

Giza, Egypt

I had the responsibility to prepare and teach two courses, namely Introduction to Programming using Python and Fundamentals of Natural Language Processing. In the meanwhile, I participated in developing TPC; a python package that do stemming, tokenization, sentence breaking, normalization, and segmentation; and also developed WebCS; a news aggregator that crawl more than 120 Arabic newswires and automatically recognize the title, body, author(s) and publish date.

Academic Background
April 2012
Master (M.Sc) in Computer Science

Cairo University

Giza, Egypt

Specialized in Web Information Retrieval and Web Personalization systems for Arabic language.
Supervisors: Dr. Kareem Darwish, QCRI/HBKU, Dr. Mervat Gheith, and Dr. Waleed Arafa, Cairo University, Egypt.

May 2006
Postgraduate Diploma in Computer Science

Cairo University, Giza, Egypt

May 2003
Bachelor of Computers and Information

Cairo University, Giza, Egypt

Majoring in: Information Systems


I have received a full scholarship from Gina Cody School of Engineering and Computer Science, Concordia University to pursue PhD in Computer Science.

I have been awarded a Concordia University International Tuition Award of Excellence, valued at approximately $37,915 .

Graduated the first of my class in the Masters program with a final grade of 82%, with a recommendation to complete the PhD.

I developed an Arabic dialect identification system using features extracted from only the transcripts of a speech recognition system. Using this system, achieved the FIRST place in accuracy and third place in F1 among 18 participants in the DSL Shared Task 2016 of the VarDial 2016 workshop for Arabic dialect identification [LINK].

A "Wikipedian"; an active Wikipedia member who have created and contributed in writing in more than 1800 articles in both Arabic and English Wikipedia projects and Obtained a scholarship to join the Wikimania Conference in Poland in 2010 [LINK]

Technical Skills