Christine Egan

Bradenton, Florida · (516)838-2494· christineegan42@gmail.com

Resourceful problem-solver with a background in management and an education in linguistics. My experiences combined with skills in data engineering, data analysis, and machine learning make me uniquely qualified to tackle the challenges of natural language processing. I specialize in extracting meaningful insights from unwieldy text and I’m driven to evolve and share my knowledge with others.

Projects

Facebook Political Ad Classification

An analysis of Facebook political ad data from Propublica using NLP with SpaCy and machine learning with SciKit-Learn.

  • Labeled 20,000 political ads as liberal or conservative using a combination of automated and manual techniques.
  • Created new features such as word embeddings with SpaCy and sentiment polarity with Vader.
  • Fit, validated, and compared three different models: Logistic Regression, Naive Bayes, and Support Vector Classifier.


Twitter and Reddit User Sentiment Analysis for Apple M1

An ETL pipeline created with Python to analyze user opinions of the Apple M1 chip.

  • Developed a CLI app to regularly collect data from Twitter and Reddit APIs.
  • Programmatically collected, cleaned, and processed batches of data, creating a data set with 13,000+ observations..
  • Engineered features with NLTK and created a mode with SciKit-Learn to predict user sentiment with 96% accuracy..

Education

Flatiron School

Certificate of Completion
Data Science Immersive Online Program
2020 - 2021

Stony Brook University

Bachelor of Arts
Lingustics and Philosophy, cum laude

Departmental Honors in Philosophy

2016 - 2019

Skills

Programming Languages & Tools
Python
Data Manipulation

  • Pandas
  • NumPy


Data Visualizations

  • Matplotlib/Seaborn
  • Plotly Dash


Tools

  • Jupyter
  • Visual Studio Code

Natural Language Processing

  • NLTK
  • SpaCy


Machine Learning

  • SciKit-Learn
  • Gensim


Webscraping

  • HTML
  • BeautifulSoup