← All work

research - In Progress

Media Framing & Sentiment Transformation Pipeline

A pipeline for converting unstructured media into structured framing and sentiment representations.

Role
Builder / Researcher
Team / Org
Solo project
Timeline
Mar 2026 – Present
Skills / Tools
  • NLP pipelines
  • LLM evaluation
  • Bias analysis
  • Structured outputs

Overview

I’m building a research-oriented transformation pipeline that combines classical NLP and LLM-based approaches to analyze sentiment, framing, and robustness across media inputs, with a focus on evaluation quality and bias awareness.

What I did

  • Building a pipeline that transforms unstructured media into structured representations.

  • Comparing classical NLP and LLM-based approaches.

  • Evaluating outputs for robustness, bias, and significance.