Research Project

Cryptocurrency Behavioral Analysis

Built with R and NLP methods.

Back to Portfolio

This research studies behavioral patterns in cryptocurrency conversations by combining statistical analysis in R with NLP workflows. The goal was to understand how sentiment, themes, and reaction timing shift as market conditions change.

Method

Collected and cleaned crypto-related text data for analysis-ready corpora.

Built NLP pipelines in R for tokenization, sentiment scoring, and topic signals.

Compared sentiment and narrative shifts across periods of market volatility.

Key Findings

Discussion sentiment moved quickly with price volatility and major news cycles.

Topic clusters exposed recurring behavioral themes during high-uncertainty windows.

NLP-based signals helped surface shifts in market narrative earlier than manual review.