Greek Life at Union College: A Century of Change

This project explores the evolution of Greek Life at Union College over 120 years (1880–2000) using archival data from the Concordiensis newspaper. By analyzing sentiment trends, linguistic patterns, and key descriptors, it provides insights into the changing perceptions of fraternities, sororities, and the broader Greek system on campus.

---

Overview

Union College has played a pivotal role in the history of Greek Life as the "Mother of Fraternities." This project leverages text data from the Concordiensis archives to examine:

• Sentiment trends toward Greek Life over time.
• Key adjectives and language shifts associated with fraternities and sororities.
• The evolution of campus narratives and culture.

By applying Natural Language Processing (NLP) techniques, including sentiment analysis and word cloud generation, this analysis uncovers how Greek Life shaped Union College's culture and vice versa.

---

Key Findings

1. Greek Life Sentiment Trends:

   • Positive sentiment surged in the late 20th century, reflecting modernization efforts.
   • Fluctuations in sentiment during the mid-20th century correspond with societal critiques and changes in campus culture.

2. Fraternity vs. Sorority Perceptions:

   • Fraternities showed consistently high positive sentiment, reflecting their deep integration into campus traditions.
   • Sororities, introduced in the 1970s, experienced predominantly positive sentiment, indicating their constructive impact on campus life.

3. Language Analysis:

   • Common descriptors such as "social," "new," and "Greek" highlight the cultural and organizational roles of these institutions.
   • Word clouds reveal the importance of terms like "brothers," "sisters," "chapter," and specific fraternity and sorority names.

---

Dataset

• Source: Concordiensis newspaper archives, Union College (1880–2000).
• Challenges: The data, extracted using Optical Character Recognition (OCR), included noise, misspellings, and character anomalies. Cleaning and preprocessing were key components of the project.

---

Methodology

1. Data Cleaning:

   • Tokenized and standardized text for analysis.
   • Extracted relevant sentences using keyword-based filtering.

2. Sentiment Analysis:

   • Leveraged VADER SentimentIntensityAnalyzer to classify positive and negative sentiments.
   • Tracked sentiment proportions yearly.

3. Visualization:

   • Generated word clouds to highlight frequently used terms.
   • Graphed sentiment trends for Greek Life and specific organizations over time.

4. Adjective Analysis:

   • Identified key descriptors by decade, providing insights into shifting narratives.

---

How to Explore the Project

1. Open the Colab notebook file (KA_Copy_of_206_Final_Project_KSHITIJ.ipynb).
2. Follow the code and comments to:
   • Understand the data cleaning and preprocessing steps.
   • Explore sentiment trends and visualizations.
   • View word clouds and adjective analyses.

Note: Ensure the necessary Python libraries (e.g., nltk, matplotlib, wordcloud) are installed if running the notebook locally.

---

Key Visuals

• Sentiment Trend Graphs: Explore the positive and negative sentiment proportions for Greek Life, fraternities, and sororities.
• Word Clouds: Visualize common themes and keywords associated with Greek organizations.

---

Acknowledgments

This project was completed as part of a final assignment for a Text Analytics class at Union College. Special thanks to Professor Nicholas Webb and the Concordiensis archives for providing the data and inspiration for this analysis.

---

Contact

For questions or feedback, feel free to reach out:

• Email: [omniscienthorizon@gmail.com]
• LinkedIn: [https://www.linkedin.com/in/coachkshitijagarwal]

---

License

This project is licensed under the MIT License. Feel free to use, modify, and share the analysis and code with proper attribution.