Data Analyst | Industrial Chemist | Data Scientist

I analyze data to solve real-world problems—turning raw numbers into actionable insights. My work spans across Business analysis, Environmental Analysis, Business operations, and visualization.

http://www.linkedin.com/in/arafahibitoye

https://github.com/heisarafat

Medium: https://medium.com/@TayoUnvrs

📁 Portfolio Overview

Welcome to my data science and analytics portfolio. Below, you’ll find some of my key projects, tools I use, certifications, and a little about my background.

⚙️ Tool Stack: SQL · Python · Excel · Numpy · Pandas · Matplotlib · Seaborn · Scikit-learn · Notion

📊 Featured Projects

1. Urhobo-English Dictionary Corpus Extraction

Goal: To extract, clean, and align translation pairs from a scanned Urhobo-English dictionary and generate a structured CSV corpus for training machine translation models. This was completed as part of a data extraction internship task for Dala Translate.

Tools: Python, pdfplumber (for PDF parsing), re (regular expressions for pattern matching), pandas (for data structuring and export), Google Colab

Highlights:

• Extracted over 599 aligned translation pairs from a noisy, unstructured PDF.
• Built a custom regex pipeline to identify and clean multi-line, multi-meaning dictionary entries.
• Output structured as a clean CSV with urhobo and english columns.
• Project can be reused for other Nigerian languages and supports building parallel corpora for NLP and machine translation research.