Amin Abdaoui

Doctor of Computer Science from the University of Montpellier, I am mainly interested in text mining applied to social networks (tweets, blogs, forums, etc.) and more generally in data science.

Currently Data Scientist NLP at AirBus BizLab and mentor for an online Data Science master's degree (OpenClassrooms).

During my thesis, I built a sentiment lexicon (polarity and emotion) for French. This lexicon is used for academic and commercial applications in France and Canada.

I also developed a sentiment analysis system that ranked first in one of the three sentiment classification tasks of the eleventh text mining challenge (DEFT 2015) and achieved second place in the latest edition DEFT 2017.

The system allows for automatic learning of the best feature and parameter configuration according to the benchmark. It has been tested on tweets, movie reviews, book reviews, video game reviews, and parliamentary debates.

I have collaborated with companies to assist them with sentiment analysis, classification, and textual data annotation tasks using brat, and finally named entity recognition.

After my thesis, I worked at the University of Montpellier for a year where I was involved in a research project in biomedical NLP in collaboration with Stanford University.

I can meet the following needs:

- Data collection and cleaning

- Data study and manual annotation (crowdsourcing..)

- Feature engineering, model training, and tuning

- Making results and models available for future use