Computational Data Analyst (Research Specialist) at the University of Minnesota Department of Political Science with experience in text analysis, web scraping, data cleaning, machine learning, natural language processing…
Computational Data Analyst (Research Specialist) at the University of Minnesota Department of Political Science with experience in text analysis, web scraping, data cleaning, machine learning, natural language processing, sentiment analysis, and data visualization. Also experienced in PostgreSQL, Power BI, Excel, TensorFlow, PyTorch, Docker, AWS, Azure, and survey design using XLSForms.
- Using Selenium to web-scrape large set of news articles from various websites
- Developing text-mining programs in Python
- Using Natural Language Processing techniques and libraries to extract and manipulate certain parts of texts
- Training a Machine Learning model for a text-classification
- Fine tuning a Chatgpt model to generate summaries and detailed analysis of news articles
- Using Excel to organize extracted information and perform various analysis
- Designing online surveys using XlsForm and advanced Excel functions and analyzing the survey results
- Managing and analyzing precipitation and temperature data of India
Performed numerous text-analysis tasks on news articles in Python to gain insights into the ongoing social conflicts in Southeast Asia.
Facilitated the process of downloading relevant news articles from multiple websites by developing an automatic web-scraping program in Python using Selenium library, which collected more than 50,000 articles in just 5 days.
Implemented robust data cleaning methodologies, including Stopword Removal, Lemmatization, and Tokenization, optimizing them for efficient Machine Analysis for Text Classification and enhancing the model’s accuracy to 94%.
Leveraged advanced Machine Learning algorithms, including Naive Bayesian and Neural Networks, to conduct binary text classifications, enabling the precise filtration of the highly relevant news articles with the 94% accuracy.
Performed comprehensive Sentiment Analysis to identify and eliminate news articles exhibiting neutral or positive sentiment, thereby intensifying the focus on articles pertaining to violence or conflict, aligning with research goals.
Utilized Natural Language Processing for each article to extract and manipulate text parts and recognize the references to the same individuals or groups for easier detection of perpetrators and victims associated with each conflict, which saved more than $1200 of research budget without using a paid AI model.
Implemented sophisticated Regular Expressions to systematically extract crucial elements, including dates, geographic locations, involved actions, and the names of perpetrators/victims, from each article.
Meticulously organized and imported the extracted information into an Excel spreadsheet using Pandas library, facilitating comprehensive analysis and streamlined data management for in-depth analysis of each social conflict.
Fine-Tuned a ChatGPT gpt-3.5-turbo model to generate concise and accurate summaries for each article.
Designed XLSForms for Caste discrimination online surveys and used conditional formatting and question branching to dynamically adjust survey questions based on users' previous responses, saving over $2000 in paper survey costs while enhancing survey efficiency and responsiveness.
Led weekly team meetings with 5 members and effectively communicated ideas for better performance of the team.
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
