Introduction
What is PTIIKInsight?
PTIIKInsight is an AI-powered topic modeling platform designed for analyzing research papers and academic content. Built specifically for JPTIIK (Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer), it provides researchers and academics with tools to discover topics and patterns in research literature.
Our Development Team
Muhammad Rajendra Alkautsar Dikna GitHub: alkadikna
Rochmanu Purnomohadi Erfitra GitHub: nrcst
I Kadek Surya Satya Dharma GitHub: Suryy16
Core Capabilities
Topic Modeling
BERTopic Integration: Uses pre-trained transformer models for accurate topic discovery
Research Paper Analysis: Specialized for academic content and research papers
Text Classification: Automatically categorize research content by topic
Batch Processing: Efficiently process multiple documents at once
Data Collection
Web Scraping: Automated collection of research papers from academic sources
Data Preprocessing: Clean and prepare text data for analysis
CSV/JSON Export: Export collected data in standard formats
Data Validation: Ensure data quality before processing
Web Interface
Streamlit Dashboard: User-friendly web interface for all operations
Real-time Predictions: Interactive topic prediction interface
Progress Monitoring: Track scraping and training operations
Results Visualization: Charts and tables for analysis results
Use Cases
Academic Research
Literature Analysis: Analyze trends in research publications
Topic Discovery: Identify emerging research areas and themes
Research Classification: Categorize papers by research domain
Content Analysis: Extract insights from academic abstracts and papers
Educational Applications
Curriculum Development: Understand current research trends for course planning
Research Guidance: Help students identify research directions
Publication Analysis: Analyze institutional research output
Knowledge Mapping: Visualize research landscape and connections
Key Features
For Researchers
Easy-to-Use Interface: No coding required for basic operations
Quick Topic Prediction: Classify new research content instantly
Data Export: Download results for further analysis
Flexible Input: Support text input, file upload, and web scraping
For System Administrators
Docker Deployment: Containerized setup for easy deployment
Monitoring Integration: Prometheus and Grafana for system health
API Access: RESTful API for programmatic access
Background Processing: Non-blocking operations for better user experience
Technology Stack
PTIIKInsight uses modern, proven technologies:
Machine Learning: BERTopic, sentence-transformers, scikit-learn
Backend API: FastAPI with async support
Web Dashboard: Streamlit with interactive components
Data Processing: pandas, numpy for data manipulation
Web Scraping: crawl4ai for automated data collection
Monitoring: Prometheus metrics, Grafana dashboards
Containerization: Docker and Docker Compose
Model Storage: Pickle format for trained models
System Requirements
Minimum Requirements:
Python 3.8 or higher
4 GB RAM
5 GB storage space
2+ CPU cores
Recommended for Production:
Python 3.10+
8 GB+ RAM
20 GB+ SSD storage
4+ CPU cores
Getting Started
Ready to start using PTIIKInsight? Follow these steps:
Installation Guide - Set up the platform (Docker or local)
Web UI Guide - Learn to use the dashboard interface
API Reference - Explore programmatic access
Docker Architecture - Understand the system design
Quick Start
Using Docker (Recommended)
git clone https://github.com/your-org/PTIIKInsight.git
cd PTIIKInsight/project
docker-compose up -dLocal Installation
cd project
pip install -r requirements.txt
pip install -r dashboard/requirements.txt
# Start API
uvicorn api.main:app --reload &
# Start Dashboard
streamlit run dashboard/main.pyAccess the dashboard at: http://localhost:8501
Support and Documentation
Installation Help: Installation Guide
Usage Instructions: Web UI Guide
API Documentation: API Reference
Troubleshooting: Troubleshooting Guide
Continue to the Installation Guide to get started with PTIIKInsight.
Last updated