This endpoint helps you discover questions that are semantically similar to your input text. It uses vector embeddings to find markets with questions that are conceptually related to your query, even if they don’t share the exact same words.

How It Works

The search process follows these steps:

  1. Your input text is converted to a vector embedding
  2. The system searches for markets with similar semantic meaning
  3. Each found question is compared to your input using cosine similarity
  4. Results are filtered and sorted by similarity score

Query Parameters

  • text (required): The input text to find related questions for
  • limit (optional): Maximum number of related questions to return (default: 5)
  • min_similarity (optional): Minimum similarity threshold between 0 and 1 (default: 0.7)

Response Format

{
  "data": [
    {
      "question": "Will US inflation exceed 3% in 2025?",
      "adj_ticker": "polymarket_will-us-inflation-exceed-3-percent-in-2025",
      "similarity": 0.89
    },
    // ... more related questions
  ]
}

Use Cases

  1. Question Discovery: Find similar questions that others have already asked
  2. Market Research: Discover related markets based on your topic of interest
  3. Content Recommendations: Suggest related questions to users

Example Query

# Find questions related to inflation
curl "https://api.data.adj.news/api/search/related-questions?text=what%20will%20happen%20to%20inflation%20rates&limit=3&min_similarity=0.8"

Notes

  • The similarity score ranges from 0 to 1, where 1 indicates perfect similarity
  • Results are sorted by similarity score in descending order
  • Only questions meeting the minimum similarity threshold are returned