> ## Documentation Index
> Fetch the complete documentation index at: https://docs.search-artemis.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Interpreting Results

> Understanding and analyzing search results from Artemis Search

## Result Structure

Each search result in Artemis Search contains the following key components:

<ResponseField name="tag" type="string">
  The content associated with the matching embedding from your dataset.
</ResponseField>

<ResponseField name="proba" type="float">
  The probability (between 0 and 1) that the result matches your search query.
</ResponseField>

## Understanding Probability Scores

The `proba` value indicates how likely it is that the result matches your search criteria:

<CardGroup cols={2}>
  <Card title="High Probability (> 0.8)" icon="circle-check">
    Strong match to your search query. These results are highly relevant.
  </Card>

  <Card title="Medium Probability (0.5 - 0.8)" icon="circle-half-stroke">
    Moderate match. May be relevant but requires closer examination.
  </Card>

  <Card title="Low Probability (< 0.5)" icon="circle-exclamation">
    Weak match. Less likely to be relevant to your search query.
  </Card>
</CardGroup>

## Analyzing Search Results

<Steps>
  <Step title="Review Top Results">
    Start by examining the results with the highest probability scores. These are most likely to be relevant to your search query.
  </Step>

  <Step title="Consider Context">
    Remember that Artemis Search understands context. A result might not contain exact keywords but could still be highly relevant due to semantic understanding.
  </Step>

  <Step title="Examine Tags">
    The `tag` field contains the actual content associated with the result. Analyze this to understand why it was returned for your query.
  </Step>

  <Step title="Adjust Parameters">
    If you're not getting the expected results, consider adjusting your [search parameters](/search/search-parameters) such as the probability threshold or synthetic dataset size.
  </Step>
</Steps>

## Example Interpretation

<CodeGroup>
  ```json theme={null}
  {
    "results": [
      {
        "tag": "HealthTech Inc: A leading provider of HIPAA-compliant healthcare software solutions.",
        "proba": 0.95
      },
      {
        "tag": "SecureData Systems: Offers data protection services for various industries.",
        "proba": 0.72
      },
      {
        "tag": "MediCare Innovations: Developing new medical devices for hospitals.",
        "proba": 0.45
      }
    ]
  }
  ```
</CodeGroup>

In this example:

1. "HealthTech Inc" is highly relevant (0.95 probability) and explicitly mentions HIPAA compliance.
2. "SecureData Systems" is moderately relevant (0.72 probability). While not specifically mentioning HIPAA, it's in a related field.
3. "MediCare Innovations" has a lower relevance (0.45 probability). It's in the medical field but doesn't indicate HIPAA compliance.

## Next Steps

* Learn more about [search parameters](/search/search-parameters)
* Understand how to [interpret search results](/search/interpreting-results)