Connecting Research to Development (CRD)Data Scientist
Sep. 2023 - Apr. 2024Dubai, United Arab EmiratesConducted data analysis project in Python, delivering insights from data collected, with findings shared with UNICEF. Utilized ArcGIS Pro to map 286 areas in Lebanon, creating a detailed grid system for building detection using deep learning features, and shared them on ArcGIS Online for facilitating the team in data collection process. Represented PHCCs in Lebanon to get how many and what cadasters each PHCC covers, and how many and what other PHCC it intersects with. Worked on extracting data from messaging platforms, utilizing cloud storage solutions for data handling. Executed advanced text analysis techniques, including topic detection, classification based on predefined topics, and sentiment analysis leveraging NLP libraries. Employed state-of-the-art machine learning models for generating topic embeddings, enhancing data consistency through cosine similarity measures. Proposed and implemented the use of some existing AI tools to facilitate communication with UNICEF, WHO, and UNFPA offices. Explored innovative OCR techniques for digitizing Arabic text from vaccination cards, initially experimenting with a custom grid system before transitioning to an advanced library for more efficient data tabulation. Applied image pre-processing enhancements to optimize text recognition
accuracy. Successfully integrated these technologies to convert unstructured data into structured data frames, significantly streamlining the data management process.