I study the consequences of conspiracy theories using NLP and computational social sciences. Explored node embedding and graph neural networks for a stance classifier, revealing patterns in social networks. Collaborated with psychology experts on cognitive biases in conspiracy beliefs and their consequences.

Business IntelligenceData ScienceBig DataSQL

Architecting an Arabic COVID-19 fake news detection system with BERT-based stance detection classifiers, creating a dashboard through social network analysis and topic modeling to detect top influencers, communities, and topics for targeted COVID-19 awareness campaigns.

Processing and analyzing big data sets, including unstructured data from social media and unconventional data sources, applying descriptive and predictive data analytics, constructing machine learning and deep learning models, focusing on natural language processing (NLP) for Arabic and low-resource dialects, and generating interactive data visualizations.

Developed a pipeline to extract job requirements from unstructured online ads, identifying the most sought-after skills in the job market using DBpedia, SBERT, and Word2Vec. Additionally, formulated a curriculum training roadmap for universities based on the analysis of 500,000 job ads, employing NLP and statistical techniques to extract top-rated professional skills and occupations.

Identified top concerns from the written contributions received by the commission from different entities. Then using social network analysis, we rank the top priorities of the different entities.