Understanding Bynder's Natural Language Search (NLS)

Summary

Natural Language Search, or NLS, helps users search for images in the portal using simple, everyday language. With NLS, users can quickly find image assets with or without tags, this makes locating the right asset faster and easier. NLS is unlike conventional searches, as it's based on high-quality image embeddings and training data to understand image content and search queries semantically.  

Who?

This feature/solution requires your Customer Success Contact to enable, but then individual permissions can be done by the Bynder Admin.

Don't yet have Bynder? Start Here!

This feature is part of Bynder's AI Search Offerings. 

How?

Searching With NLS

  1. Click the search bar.
  2. A  banner will appear.
  3. Click the toggle to turn NLS on. Once activated, the toggle will remain on until manually turned off.
  4. Type in a phrase to describe the asset needed.
  5. The asset overview will display the matching results.

Tip: Users can combine search methods and use NLS alongside filters or keyword-based searches to further refine results.

Best Practices

  • Use natural, everyday terms. 
  • Be precise and clear.
  • Avoid vague terms, slang, or esoteric wording.
  • Avoid time and location specific prompts unless the metadata has been compiled with this information. 

Supported Languages

Bynder’s Natural Language Search is fully optimized for English. While also supporting simple queries in languages such as Spanish, Portuguese, Catalan, Italian, German, Dutch, and Polish, the most accurate and reliable results are achieved in English.

FAQs

How many search results does NLS produce?
NLS can provide a max of 500 results, which are then ranked by the AI model.

Can I apply a filter first and then search using NLS?
Yes you can, and both the search and the filter will be applied to your results.

Does NLS support conversational language?
No, it currently does not support conversational language. 

How does NLS search?
NLS uses high-quality image embeddings and training data to understand image content and search queries semantically. It does not use metadata and may struggle with highly specific searches where metadata excels. Review what works best for your specific query.

Can I apply a ‘search by image’ or a ‘similarity search’ and then search using NLS?
You cannot combine these; the latest search action will override the former.

Related Articles

Bynder's AI Search Experience Offerings

Understanding Universal Compact View (UCV)

Updated