Mark Ma.
← All work

Thesis · NLP & data engineering

Twitter Sentiment Analysis

A thesis on public sentiment around COVID protective measures, supported by a complete analytics pipeline.

Academic research · Open-source code

The question

How did public sentiment about vaccination, masks, distancing, working from home, and sanitizer relate to state-level COVID case trends? This thesis investigated those associations using public Twitter posts.

The work behind the analysis

Collected millions of tweets and built an analytics pipeline using Kafka, Google Cloud, and BERT. The project combined data collection and engineering with sentiment analysis and interpretation.

Reading the result carefully

The analysis explores correlations in online discussion, rather than establishing that sentiment or a protective measure caused a change in case counts.

Explore the code ↗
← Back to selected workGet in touch ↗