NarrativeNexus is an end-to-end NLP-powered web application built using Streamlit that performs advanced text analysis on multiple document formats. It integrates text preprocessing, sentiment analysis, topic modeling, summarization, visualization, and automated reporting into a single interactive platform.
- Frontend/UI: Streamlit
- Backend: Python
- NLP Libraries: NLTK, Scikit-learn
- Deep Learning: Hugging Face Transformers
- Visualization: Matplotlib, WordCloud
- Reporting: ReportLab
- Data Handling: Pandas
NarrativeNexus/
│
├── app.py # Entry point
├── streamlit_ui.py # UI rendering logic
├── collection.py # File handling & text extraction
├── preprocessing.py # Text preprocessing pipeline
├── cleaning.py # Regex-based cleaning
├── sentiment.py # Sentiment analysis (VADER)
├── model.py # LDA topic modeling
├── summarization.py # Extractive + Abstractive summarization
├── insights.py # Insight generation
├── visualization.py # Word cloud & charts
├── reporting.py # PDF report generation
├── css.py # Custom UI styling
└── requirements.txtInput (Files/Text)
↓
Text Extraction
↓
Preprocessing (Clean → Tokenize → Lemmatize)
↓
Sentiment Analysis
↓
Topic Modeling (LDA)
↓
Summarization (Extractive + Abstractive)
↓
Insight Generation
↓
Visualization Dashboard
↓
Report Generation (PDF + CSV)1️⃣ Clone Repository
git clone https://github.com/Keerthana-webdev/Narrative-Nexus.git
cd NarrativeNexus
2️⃣ Install Dependencies
pip install -r
requirements.txt
3️⃣ Run Application
streamlit run app.py
- Upload files or paste text
- Click "🚀 Analyze Text"
- Explore:
- Text statistics
- Sentiment results
- Topics
- Summaries
- Insights & visualizations
- Download:
- CSV output
- PDF report
- ✔ Sentiment classification with scores
- ✔ Topic keywords (LDA)
- ✔ Extractive & abstractive summaries
- ✔ Word cloud visualization
- ✔ Sentiment distribution charts
- ✔ Auto-generated PDF report
- Real-time API integration
- Multi-language support
- Named Entity Recognition (NER)
- Keyword trend analysis
NarrativeNexus demonstrates a complete NLP pipeline by combining classical machine learning and modern deep learning techniques into a scalable, user-friendly platform. It bridges the gap between raw text data and actionable insights.