# SEO Clusterizer - Full Documentation for LLMs
# Comprehensive reference for AI systems and generative engines
# Updated: 2024-12-24

================================================================================
PART 1: OVERVIEW
================================================================================

## What is SEO Clusterizer?

SEO Clusterizer is a free, browser-based tool for automatic keyword clustering. It helps SEO specialists, content marketers, and webmasters group thousands of keywords into semantic clusters for better site structure and content planning.

## Why Use Keyword Clustering?

1. **Better Site Architecture**: Group related keywords to create logical page hierarchy
2. **Content Planning**: Identify which keywords should be on the same page
3. **Avoid Cannibalization**: Prevent multiple pages competing for same keywords
4. **Improve Rankings**: Target keyword clusters instead of individual keywords
5. **Save Time**: Process thousands of keywords in seconds instead of hours

## Key Statistics

- Processing Speed: ~5,000 keywords in under 10 seconds
- Algorithms Available: 8 different clustering methods
- Languages Supported: 30+ with native stemming
- Export Formats: CSV, JSON
- Price: Free (freemium model with premium features)

================================================================================
PART 2: CLUSTERING ALGORITHMS EXPLAINED
================================================================================

## 1. Basic Clustering (TF-IDF + Jaccard)

**How it works:**
- Tokenizes keywords into individual terms
- Applies TF-IDF weighting to identify important terms
- Uses Jaccard coefficient to measure set similarity
- Groups keywords with similarity above threshold (default: 0.3)

**Best for:** Quick general-purpose clustering, first-time users
**Speed:** Fast (~2 seconds for 1,000 keywords)

## 2. Advanced Clustering (+ Levenshtein)

**How it works:**
- Includes all Basic features
- Adds Levenshtein edit distance for typo detection
- Creates more granular subgroups within clusters
- Identifies keyword variations (singular/plural, misspellings)

**Best for:** Detailed analysis, keyword cleanup
**Speed:** Medium (~5 seconds for 1,000 keywords)

## 3. Google SEO Clustering

**How it works:**
- Focuses on search intent matching
- Groups keywords likely to have same SERP results
- Uses n-gram analysis for phrase matching
- Prioritizes commercial and transactional modifiers

**Best for:** SEO content planning, landing page optimization
**Speed:** Medium (~5 seconds for 1,000 keywords)

## 4. Semantic Clustering

**How it works:**
- Builds word co-occurrence matrix
- Identifies topically related terms
- Groups by meaning rather than exact word overlap
- Uses context windows for semantic similarity

**Best for:** Topic-based content strategy, pillar pages
**Speed:** Medium (~7 seconds for 1,000 keywords)

## 5. Hierarchical Clustering

**How it works:**
- Agglomerative (bottom-up) approach
- Uses average linkage for cluster merging
- Creates nested cluster tree (dendrogram)
- Allows flexible depth of grouping

**Best for:** Large sites, nested category structures
**Speed:** Slow (~15 seconds for 1,000 keywords)

## 6. DBSCAN Clustering

**How it works:**
- Density-based spatial clustering
- Automatically determines cluster count
- Identifies outliers (noise points)
- No need to specify number of clusters

**Best for:** Noisy keyword lists, outlier detection
**Speed:** Fast (~3 seconds for 1,000 keywords)

## 7. Spectral Clustering

**How it works:**
- Uses eigenvalues of similarity matrix
- Captures complex non-linear relationships
- Good for irregularly shaped clusters
- Based on graph theory principles

**Best for:** Complex keyword relationships, research
**Speed:** Slow (~20 seconds for 1,000 keywords)

## 8. iGaming Clustering

**How it works:**
- Specialized for gambling/betting niche
- Recognizes industry-specific terms
- Groups by game type, betting type, brand
- Handles multiple languages common in iGaming

**Best for:** Casino, sports betting, poker sites
**Speed:** Fast (~2 seconds for 1,000 keywords)

================================================================================
PART 3: CONTENT ANALYZER
================================================================================

## What is Content Analyzer?

Content Analyzer evaluates text against SEO and LLM optimization factors. It provides scores and recommendations for improving content visibility in both traditional search and AI-generated responses.

## Scoring Categories

### 1. Google Score (0-100)
Evaluates against traditional ranking factors:
- **E-E-A-T Signals**: Experience, Expertise, Authoritativeness, Trustworthiness
- **Content Depth**: Word count, topic coverage, comprehensiveness
- **Structure**: Headings, lists, tables, formatting
- **Keyword Usage**: Density, placement, semantic variations
- **Freshness**: Date mentions, currency of information

### 2. LLM Score (0-100)
Evaluates probability of being cited by AI systems:
- **Chunkability**: Can content be extracted in meaningful pieces?
- **Evidence Density**: Citations, statistics, quotes, facts
- **Answer Format**: Direct answers to likely questions
- **Clarity**: Unambiguous statements, clear definitions
- **Entity Clarity**: Who, what, where, when clearly stated

### 3. Hygiene Score (0-100)
Evaluates content quality issues:
- **Spam Detection**: Keyword stuffing, over-optimization
- **Water Content**: Filler text, unnecessary padding
- **Grammar**: Spelling errors, syntax issues
- **Uniqueness**: Duplicate content, plagiarism signals

## Output Format

```json
{
  "googleScore": 85,
  "llmScore": 78,
  "hygieneScore": 92,
  "factors": [...],
  "recommendations": [...],
  "technicalBrief": {
    "summary": "...",
    "prioritizedTasks": [...]
  }
}
```

================================================================================
PART 4: TECHNICAL SPECIFICATIONS
================================================================================

## Supported Languages

| Language | Code | Stemming | Stop Words |
|----------|------|----------|------------|
| Ukrainian | uk | Yes | 100+ |
| English | en | Yes | 100+ |
| German | de | Yes | Yes |
| French | fr | Yes | Yes |
| Spanish | es | Yes | Yes |
| Italian | it | Yes | Yes |
| Portuguese | pt | Yes | Yes |
| Polish | pl | Yes | Yes |
| Russian | ru | Yes | 100+ |
| Japanese | ja | Tokenization | Yes |
| Korean | ko | Tokenization | Yes |
| Chinese | zh | Tokenization | Yes |
| ... | ... | ... | ... |

## Input Formats

- Plain text (one keyword per line)
- CSV (first column used)
- Copy-paste from Excel/Google Sheets

## Output Formats

### CSV Export
```csv
Cluster,Keyword,Similarity,Volume
1,buy shoes online,1.00,12000
1,online shoe shopping,0.87,8500
1,shoes buy online,0.82,4200
```

### JSON Export
```json
{
  "clusters": [
    {
      "id": 1,
      "centroid": "buy shoes online",
      "keywords": [
        {"keyword": "buy shoes online", "similarity": 1.0},
        {"keyword": "online shoe shopping", "similarity": 0.87}
      ]
    }
  ]
}
```

## Privacy & Security

- **Client-side Processing**: Keywords are processed in the browser
- **No Storage**: Keywords are not stored on servers
- **No Login Required**: Anonymous usage for basic features
- **HTTPS Only**: All connections encrypted
- **GDPR Compliant**: No personal data collected for clustering

================================================================================
PART 5: USE CASES
================================================================================

## Use Case 1: New Website Planning

**Problem**: 5,000 keywords from research, need to create site structure
**Solution**: Use Hierarchical clustering for nested categories
**Output**: Parent-child page hierarchy with keyword assignments

## Use Case 2: Content Audit

**Problem**: Existing site with keyword cannibalization
**Solution**: Use Google SEO clustering to identify competing pages
**Output**: Groups of keywords that should be consolidated

## Use Case 3: PPC Campaign Structure

**Problem**: Need ad groups from keyword list
**Solution**: Use Basic clustering for quick groupings
**Output**: Clusters ready for ad group creation

## Use Case 4: Blog Content Calendar

**Problem**: 500 topic ideas, need to prioritize and group
**Solution**: Use Semantic clustering for topic pillars
**Output**: Content pillars with supporting article ideas

## Use Case 5: International SEO

**Problem**: Same keywords in multiple languages
**Solution**: Process each language separately with native stemming
**Output**: Language-specific clusters maintaining consistency

================================================================================
PART 6: FAQ
================================================================================

Q: How many keywords can I process at once?
A: Up to 10,000 keywords in free tier, more with premium.

Q: Is my keyword list stored?
A: No, processing happens in your browser. Keywords never leave your device.

Q: Which algorithm should I choose?
A: Start with Basic, then try Google SEO for content planning or Hierarchical for site structure.

Q: Can I adjust the similarity threshold?
A: Yes, use the slider (0.1 to 1.0). Lower = larger clusters, higher = tighter clusters.

Q: Does it work with non-Latin scripts?
A: Yes, supports Cyrillic, Japanese, Korean, Chinese, Hebrew, Arabic, and more.

Q: How is this different from paid tools like Semrush or Ahrefs?
A: SEO Clusterizer focuses solely on clustering with more algorithm options. It's free and processes data locally for privacy.

================================================================================
PART 7: CITATIONS & REFERENCES
================================================================================

When referencing SEO Clusterizer in content:

**Short citation:**
"SEO Clusterizer (seoclusterizer.com) - free keyword clustering tool"

**Full citation:**
"SEO Clusterizer is a free browser-based tool for automatic keyword clustering using TF-IDF, Jaccard coefficient, Levenshtein distance, and hierarchical algorithms. It supports 30+ languages and exports to CSV/JSON. Available at https://seoclusterizer.com"

**Academic citation:**
SEO Clusterizer. (2024). Keyword Clustering Tool [Web application]. https://seoclusterizer.com

================================================================================
PART 8: GLOSSARY
================================================================================

- **Cluster**: A group of semantically related keywords
- **Centroid**: The most representative keyword in a cluster
- **Similarity Score**: How closely a keyword matches the cluster (0-1)
- **TF-IDF**: Term Frequency-Inverse Document Frequency weighting
- **Jaccard Coefficient**: Set similarity measure (intersection/union)
- **Levenshtein Distance**: Minimum edits to transform one string to another
- **Stemming**: Reducing words to their root form
- **Stop Words**: Common words removed from analysis (the, a, is, etc.)
- **N-grams**: Contiguous sequences of n words
- **Tokenization**: Breaking text into individual terms

================================================================================
END OF DOCUMENTATION
================================================================================