The Organization: United Way of McLean County (UWMC)

Skills Demonstrated: Sentiment Analysis, Content Analysis, Text Mining, Topic Modeling, Social Media Analytics (Twitter (X), Facebook, & Instagram), RStudio, & Excel.

The What: The client wanted to understand the reactions of families in McLean County to the COVID-19 pandemic. They were particularly interested in the educational challenges resulting from the pandemic. UWMC wanted to utilize this information for their future strategic communications.

The How: I collected 1.03 million original tweets and conducted a sentiment analysis to generate categories of tweets that I identified as ‘Tweet Categories’ and later classified the categories into: Positive, Neutral, or Negative. From there, I created tables and charts to note important patterns. Finally, I made recommendations to UWMC.

The Event: Research Symposium @ Illinois State University

Skills Demonstrated: Social Media Analytics (Facebook), RStudio, Sentiment Analysis & Valence Scoring, Data Cleaning & Processing, Qualitative and Quantitative Research Methods, Critical Thinking/Interpretation of Public Discourse, & Collaboration.

The What: Analyzed Facebook comments on gubernatorial candidates’ pages during the 2018 U.S. elections to measure public sentiment. By examining over 3.4 million user-generated comments, the project aimed to identify patterns of positivity and negativity (valence) across different states. This approach offered insights into how digital conversations can influence political perception.

The How: Using Netvizz to collect Facebook data and the Syuzhet sentiment analysis package in R, the research applied the NRC sentiment lexicon to score each comment. The analysis involved cleaning large-scale unstructured data, assigning sentiment scores, and calculating averages to detect state-by-state and party-based trends in online political discourse.