How bilall2003/social-media-sentimental-analysis works
This is a Streamlit dashboard for exploring social-media sentiment data and analyzing entered text. Users can inspect dataset summaries, explore vocabulary and filtered posts, and compare machine-learning and VADER sentiment results alongside NMF topic keywords and dataset-wide topic context. The runtime is concentrated in script/main.py; the README documents capabilities whose implementation details are not visible in the excerpts, so uncertain internal wiring is omitted. The tree contains training and test CSVs; their loading path is not shown.
Main entry points: main.py Streamlit app
Components 14
Dashboard & Exploration
- Streamlit app
script/main.py - Data overview
script/main.py - EDA visualizations
script/main.py - Vocabulary search
script/main.py - Text filtering
script/main.py
Sentiment Analysis
- TF-IDF features
script/main.py - ML sentiment model
script/main.py - VADER engine
script/main.py - Sentiment results
script/main.py
Topic Discovery
- NMF topic model
script/main.py - Topic keywords
script/main.py
Dataset
- Training data
Data/train.csv - Test data
Data/test.csv
Other components
- Dashboard user
Connections 13
- Dashboard user → Streamlit app: uses
- Streamlit app → Data overview: shows
- Streamlit app → EDA visualizations: shows
- Streamlit app → Vocabulary search: offers
- Streamlit app → Text filtering: offers
- Streamlit app → ML sentiment model: invokes
- Streamlit app → VADER engine: invokes
- Streamlit app → NMF topic model: invokes
- ML sentiment model → TF-IDF features: transforms text
- Vocabulary search → TF-IDF features: uses features
- ML sentiment model → Sentiment results: produces
- VADER engine → Sentiment results: produces scores
- NMF topic model → Topic keywords: produces