discourser

A Streamlit application for analyzing the influence of a core corpus of materials against another corpus.

• Python 3.10

Installation

1. Clone this repository
2. Create a virtual environment: python -m venv venv
3. Activate virtual environment: source venv/bin/activate (Linux/Mac) or venv\Scripts\activate (Windows)
4. Install dependencies: pip install -r requirements.txt

You can create a conda environment with:

conda create -n discourser python=3.10
conda activate discourser

Running the Application

streamlit run app.py

Usage

1. Project Setup: Create or load a project
2. Core Corpus: Upload CSV metadata + text files for influential texts
3. Target Corpus: Upload corpus to analyze for influence
4. Analysis:
5. Results:

Nb the 'save project' is just saving your core and target corpuses; you still need to initialize the embeddings and so on.

Data Format

CSV Metadata

Required columns: filename, title, author, date, source Optional columns: document_type

Text Files

• UTF-8 encoded .txt files
• Paragraphs separated by double blank lines
• Filename must match CSV metadata

graph TD
    A[Project Setup] --> B[Core Corpus Upload]
    A --> C[Target Corpus Upload]
    
    B --> D[Core Corpus Processing]
    C --> E[Target Corpus Processing]
    
    D --> F[Core Embeddings Generation]
    E --> G[Target Embeddings Generation]
    
    F --> H[Term Extraction from Core]
    H --> I[Suggested Terms for Vectors]
    
    F --> J[Core-Target Similarity Matrix]
    G --> J
    
    J --> K[Most Influential Core Texts]
    
    I --> L[Custom Vector Creation]
    F --> L
    
    L --> M[Vector Projection Analysis]
    F --> M
    G --> M
    
    M --> N[2D/3D Vector Spaces]
    
    F --> O[Topic Modeling on Core]
    O --> P[Topic-Target Analysis]
    G --> P
    
    J --> Q[Results & Visualization]
    K --> Q
    N --> Q
    P --> Q
    
    subgraph "Core Corpus Flow"
        B
        D
        F
        H
        I
        O
    end
    
    subgraph "Target Corpus Flow"
        C
        E
        G
    end
    
    subgraph "Comparative Analysis"
        J
        K
        M
        N
        P
    end
    
    subgraph "Vector Analysis"
        L
        M
        N
    end
    
    style B fill:#e1f5fe
    style C fill:#f3e5f5
    style F fill:#e8f5e8
    style G fill:#fff3e0
    style J fill:#ffebee
    style L fill:#f1f8e9
    style O fill:#e0f2f1

Loading