How cabooseexia/dsa4264-problem-2 works
This repository is a Reddit hate-speech detection and analysis project, combining interactive data preparation and labeling, model experimentation, topic and behavioral analysis, and an explanatory dashboard. The documented data workflow begins with Reddit threads and labeled examples, then develops analysis outputs that the dashboard presents through filters, trends, post summaries, and model explanations. The tree and README identify broader notebook capabilities, but their internal calls and data handoffs were not sampled; those links are therefore omitted rather than inferred. Dashboard-to-shared helper usage and the two interactive labeling scripts are visible in the excerpts.
Main entry points: app.py Shiny Analysis Dashboard · setup_for_expertly_labelled_data.py Expert Sample Preparation · expert_data_labeller.py Expert Labeling · Reddit_Report4.ipynb Analysis Reporting
Components 17
Data Preparation & Labeling
- Expert Sample Preparation
for_expertly_labelling/setup_for_expertly_labelled_data.py - Expert Labeling
for_expertly_labelling/expert_data_labeller.py - Data Labeling
phase_1/Labelling.ipynb
Modeling & Topic Analysis
- Model Experiments
phase_1/BERT_and_Llama_model_testing.ipynb - API Model Experiments
phase_1/api_model_testing.ipynb - Reddit BERT Analysis
phase_2/Reddit_BERT.ipynb - Post Title Extraction
phase_2/Reddit_Post_Title_Extraction2.ipynb - Post Topic Processing
phase_2/Reddit_Data_Processing_&_Post_Topics2.ipynb - Analysis Reporting
phase_2/Reddit_Report4.ipynb
Interactive Dashboard
- Shiny Analysis Dashboard
dashboard-reddit/app.py - Model & Explanation Helpers
dashboard-reddit/shared.py
Behavior & Explainability
- Topic Time Series
phase_3/TopicModelling_TimeSeries.ipynb - User Behavior Analysis
phase_3/User_Behaviour.ipynb - Model Explainability
phase_4/XAI.ipynb
Other components
- Analyst / Labeler
- Reddit Thread Data
- Dashboard User
Connections 16
- Reddit Thread Data → Expert Sample Preparation: samples and cleans
for_expertly_labelling/setup_for_expertly_labelled_data.py - Expert Sample Preparation → Analyst / Labeler: requests selection
for_expertly_labelling/setup_for_expertly_labelled_data.py - Expert Sample Preparation → Expert Labeling: saves selected rows
- Analyst / Labeler → Expert Labeling: supplies labels
for_expertly_labelling/expert_data_labeller.py - Expert Labeling → Reddit Thread Data: writes labeled CSV
for_expertly_labelling/expert_data_labeller.py - Dashboard User → Shiny Analysis Dashboard: filters and explores
- Shiny Analysis Dashboard → Model & Explanation Helpers: uses model helpers
dashboard-reddit/app.py - Model & Explanation Helpers → Shiny Analysis Dashboard: returns predictions and explanations
dashboard-reddit/shared.py - Data Labeling → Model Experiments: informs experiments
- Model Experiments → Reddit BERT Analysis: supports Reddit modeling
- Post Title Extraction → Post Topic Processing: feeds topic analysis
- Post Topic Processing → Topic Time Series: enables trend analysis
- Reddit BERT Analysis → Analysis Reporting: informs reporting
- Topic Time Series → Analysis Reporting: informs reporting
- User Behavior Analysis → Analysis Reporting: informs reporting
- Model Explainability → Shiny Analysis Dashboard: provides explainability capability