What is the "Percent Match" score in my search results, and how should I interpret it?

Percent Match is a single, easy-to-read metric (expressed as a percentage) that indicates how closely the returned document aligns with what you entered in your search. Depending on whether you used Similarity Search or Concept Search, it measures slightly different factors:

1. Similarity Search

   • The score reflects how closely the text in the result matches the text you pasted in the Text to Compare field.

   • A higher value (e.g., 80% or 90%) means the language in the result is very similar to your source text, while a lower value means there’s less overlap in wording or phrasing.

2. Concept Search

   • The score indicates how many of your required concepts appear in the returned document.

   • A higher value means the document covers more of the ideas you listed (or were extracted from your uploaded files).

• High Score:

  • Similarity Search – The result’s wording is close to your original text.

  • Concept Search – The document includes most (or all) of your specified concepts.

• Moderate Score:

  • The document shares some language or concepts but may not be a comprehensive match.

• Low Score:

  • Similarity Search – The text is only tangentially related or uses very different wording.

  • Concept Search – The document might address only one or two of your listed ideas.

Tip: A high Percent Match doesn’t necessarily mean the result addresses your specific research purpose, which is why Prompt Match can be True or False. For instance, a document can share a lot of wording or concepts but still not align with your User Prompt focus, resulting in a high Percent Match but a False Prompt Match.