Project Goals and Scope
Relevant source files
The following files were used as context for generating this wiki page:
The primary goal of the "Analisis Sentiment Data Mining" project is to develop an advanced, real-time AI-powered dashboard for sentiment analysis of YouTube comments, with a specific focus on Indonesian football (Timnas). The system is designed to move beyond simple sentiment classification by providing truly meaningful insights through intelligent comment filtering, context-aware AI models, and a high-performance, interactive user interface.
The project's scope encompasses the entire data pipeline, from real-time data fetching from the YouTube API to processing, AI-driven analysis, and visualization in a modern Next.js dashboard. A key architectural principle is the use of a dual-model system to optimize for different use cases: a pre-trained SVM model for historical data analysis and the Google Gemini AI for live, contextual analysis. The ultimate aim is to provide actionable intelligence for media analysis, social listening, and research.
Sources: README.md, DEPLOYMENT_SUMMARY.md, dashboard-next/README.md
Core Objectives and Key Features
The project is centered around delivering a set of core features that define its scope and capabilities. These features aim to solve the problem of low-quality, irrelevant comments and simplistic sentiment analysis that often fails to capture true public opinion.
Sources: README.md:5-11, DEPLOYMENT_SUMMARY.md:12-21
| Feature | Description |
|---|---|
| Aggressive AI Sentiment Analysis | An AI model designed to minimize ambiguous "neutral" classifications, reducing them by up to 96% to provide clearer positive/negative sentiment signals. |
| Smart Comment Filter | An AI-powered system using Google Gemini to filter out spam, promotions, and irrelevant content, ensuring that only comments related to football and the national team are analyzed. This improves the relevance rate from ~40% to over 90%. |
| Real-time Analysis | The capability to fetch and analyze YouTube comments live, providing up-to-the-minute sentiment monitoring during or after matches. |
| Indonesian Optimized | The models are specifically tuned to understand Indonesian slang, local expressions, negation, and context, which is crucial for accurate analysis of the target content. |
| Interactive Dashboard | A modern, responsive web application built with Next.js 14, featuring interactive charts, data visualizations, and a structured layout for exploring the analysis results. |
Sources: README.md:5-11, DEPLOYMENT_SUMMARY.md:12-21, 30, CHANGELOG.md:5-10
Smart Comment Filtering Process
The Smart Comment Filter is a critical component that narrows the scope of analysis to relevant data. It uses a scoring system to determine comment relevance.
This diagram illustrates the filtering logic applied to incoming comments.
Architectural Goals
A fundamental goal of the project is to implement a flexible and efficient dual-model architecture. This design choice allows the system to use the best tool for each specific task, balancing speed, cost, and accuracy.
Sources: README.md:13-37, CHANGELOG.md:12-26, dashboard-next/README.md:5-20
This diagram shows the dual-model architecture.
Model Comparison
| Feature | SVM Model (Static) | Gemini AI (Live) |
|---|---|---|
| Data Source | Pre-labeled & cleaned | Real-time & raw |
| Accuracy | 87.5% (domain-specific) | 95%+ (contextual) |
| Speed | <1ms per comment | 1-2s per comment |
| Context Handling | Limited | Excellent |
| Cost | Free | API quota-based |
| Reasoning | No | Yes |
Sources: README.md:29-37, CHANGELOG.md:12-26
Scope of Analysis: 5-Layer Framework
The project's analysis scope is defined by an "Optimized 5-Layer Sentiment Framework". This framework ensures that the analysis goes beyond a simple positive/negative score to provide deep, actionable insights by identifying who is being criticized, why they are being criticized, and the nature of the feedback.
Sources: config/sentiment_config_v2_optimized.py:1-11, 281-301
This diagram outlines the five layers of the sentiment analysis framework.
| Layer | Purpose | Description |
|---|---|---|
| 1. Core Sentiment | Classify the basic emotion (WHAT). | Determines if the sentiment is positive, negative, or neutral. |
| 2. Target Kritik | Identify the entity being blamed (WHO). | Categories include pssi_management, coaching_staff, and players. |
| 3. Root Cause | Diagnose the reason for failure (WHY). | Identifies issues in tactical_strategic, technical_skill, mental_psychological, or systemic_structural areas. |
| 4. Time Perspective | Determine the solution timeline (WHEN). | Classifies suggestions as immediate, short_term, or long_term. |
| 5. Constructiveness | Evaluate feedback quality (HOW). | Differentiates between constructive feedback with solutions, destructive complaints, and hopeful remarks. |
Sources: config/sentiment_config_v2_optimized.py:13-278, 294-299
Dashboard and UI Scope
The project's scope includes a comprehensive and well-structured user interface to present the analysis results. The dashboard is organized into logical sections to facilitate "Progressive Disclosure," allowing users to start with a high-level overview and drill down into details as needed.
Sources: DASHBOARD_STRUCTURE.md, MENU_GUIDE.md
This diagram shows the main navigation structure of the dashboard.
The user journey is designed for different personas:
- Casual User: Views the main Dashboard for a quick overview.
- Analyst: Navigates from the Dashboard to Sentiment and Emotion pages for deeper analysis.
- Researcher/Developer: Utilizes all pages, including the Comments Explorer and Model Performance sections.
Sources: DASHBOARD_STRUCTURE.md:112-119, MENU_GUIDE.md:204-215
Performance and Reliability Goals
A key part of the project's scope is to meet specific performance and reliability targets. These goals ensure a fast, smooth, and dependable user experience.
| Category | Metric | Target |
|---|---|---|
| Performance | Lighthouse Score | > 90 |
| API Response Time | < 100ms (cached: < 5ms) | |
| Page Load Time | < 2s | |
| Bundle Size | < 500KB | |
| Reliability | Uptime | > 99.9% |
| Error Rate | < 0.1% |
Sources: DEPLOYMENT_CHECKLIST.md:121-134, PERFORMANCE_OPTIMIZATION_REPORT.md:20-25, 93-99
These goals are achieved through a series of optimizations, including server-side caching, client-side data fetching strategies with SWR, React component memoization, and dynamic imports (code splitting).
Sources: PERFORMANCE_OPTIMIZATION_REPORT.md:28-85
Conclusion
The project's goals and scope are ambitious yet well-defined, aiming to create a state-of-the-art sentiment analysis platform for a specific, high-impact domain. By integrating a dual-model AI architecture, a sophisticated 5-layer analysis framework, and a high-performance interactive dashboard, the project is scoped to deliver far more than basic sentiment scores. It is designed to produce actionable intelligence, identify root causes of public sentiment, and provide a valuable tool for stakeholders in the Indonesian football community. The clear performance targets and structured UI design further underscore the commitment to delivering a professional and reliable final product.
Sources: README.md, DEPLOYMENT_SUMMARY.md, config/sentiment_config_v2_optimized.py