Semantic Clustering

Function Details

Outputs

This node outputs a table with cluster assignments. You can choose different types of cluster labels:

Basic numbers (integers)
Short meaningful labels
Cluster descriptions

Label Types

You can select one or more of these label types:

Basic (numbers): Simple numeric cluster assignments (0, 1, 2, etc.)
Short Meaningful Labels: Brief 2-4 word descriptions of each cluster
Cluster Descriptions: One-sentence explanations of what makes each cluster unique

Examples

Given a table of customer data:

[
  { "age": 31, "income": 30000, "purchases": 12 },
  { "age": 42, "income": 50000, "purchases": 24 },
  { "age": 25, "income": 25000, "purchases": 8 }
]

With Basic Numbers Only:

[
  { "age": 31, "income": 30000, "purchases": 12, "cluster": 0 },
  { "age": 42, "income": 50000, "purchases": 24, "cluster": 1 },
  { "age": 25, "income": 25000, "purchases": 8, "cluster": 0 }
]

With Short Labels:

[
  { "age": 31, "income": 30000, "purchases": 12, "cluster_label": "Young Moderate Spenders" },
  { "age": 42, "income": 50000, "purchases": 24, "cluster_label": "Established High Spenders" },
  { "age": 25, "income": 25000, "purchases": 8, "cluster_label": "Young Moderate Spenders" }
]

With All Label Types:

[
  {
    "age": 31,
    "income": 30000,
    "purchases": 12,
    "cluster": 0,
    "cluster_label": "Young Moderate Spenders",
    "cluster_description": "Younger customers with moderate income and average purchase frequency"
  },
  {
    "age": 42,
    "income": 50000,
    "purchases": 24,
    "cluster": 1,
    "cluster_label": "Established High Spenders",
    "cluster_description": "Middle-aged customers with higher income and frequent purchases"
  },
  {
    "age": 25,
    "income": 25000,
    "purchases": 8,
    "cluster": 0,
    "cluster_label": "Young Moderate Spenders",
    "cluster_description": "Younger customers with moderate income and average purchase frequency"
  }
]

Configuration

Set the number of clusters using the num_clusters parameter
Choose your preferred label types:
Select "Basic (numbers)" for simple numeric clusters
Select "Short Meaningful Labels" for brief descriptive labels
Select "Cluster Descriptions" for detailed explanations
Select multiple options to include multiple label types

The node will automatically generate meaningful labels and descriptions using AI when those options are selected.