> For the complete documentation index, see [llms.txt](https://raic-labs.gitbook.io/raic-fed/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://raic-labs.gitbook.io/raic-fed/analyst-guide/03d-reading-your-search-results.md).
# Reading your search results
When a search completes, RAIC Fed returns results as a context map: a visual display of image chips clustered by semantic similarity in a shared vector space. Results closest to the reference image appear in red and orange; results that are more visually different appear in yellow, green, and blue.
This view is designed for speed. The Analyst does not need to review every result individually — the clustering shows where similar objects are concentrated, making it possible to assess a large result set at a glance and act on what matters.
***
## Understanding the context map
At the top of the screen, the **Search Scope** bar summarizes the search that produced the results:
* Date range covered
* Number of scenes searched
* Detector used
The context map displays each detection as an image chip. Chips are positioned by visual similarity: objects that look alike cluster together, regardless of when or where they were found in the dataset. The color scale in the bottom right corner indicates similarity, from **Similar** (red/orange) to **Different** (blue).
> **Note** Results are not filtered or ranked by location or time. The clustering is purely visual, based on what the objects look like relative to the reference image.
***
## Selecting images from the context map
Before organizing results or inspecting them in map context, the Analyst selects the image chips to work with. Two selection tools are available in the top right toolbar of the context map:
* **Select individual images** — the top icon (square with finger). Select this to click individual chips one at a time.
* **Draw polygon** — the polygon icon. Select this to draw one or more polygons around a group of chips to select them all at once.
> **Analyst** Precision is not required at this stage. Selecting a broad group and refining within the taxonomy is faster than being exact in the context map. This workflow is built for iteration.
***
## Organizing results into a taxonomy
The left panel shows the current taxonomy categories in the shared workspace. The taxonomy is a user-defined, evolving library that turns raw search results into structured intelligence.
> **Note** The taxonomy is shared across all Analysts in the workspace. Every decision an Analyst makes adds to it, making every future search more precise.
### Saving images to an existing category
1. Select images from the context map using the selection tools described above.
2. In the left panel, select the category to assign the images to.
3. Select **Save Images to Category** to confirm.
### Adding a new category
Categories can be nested. To add a new category at any level:
1. In the left panel, select the category level to add the new category to. The selected level will highlight.
2. Select **Add New Category** and name it.
3. Select images from the context map and save them to the new category.
***
## Inspecting results in map context
Once images are selected, the Analyst can expand the right panel to see the selected detections in their original map context, including the exact location in the source imagery where each object was found.
1. With images selected, select the **left arrow** on the right edge of the screen to expand the map panel.
2. The selected detection is shown highlighted in the full source imagery on the right, with date, sensor, and coordinates displayed below.
3. Choose a grouping on the map to display filmstrip of images on bottom right of the page.
4. Use the filmstrip at the bottom to scroll through the selected detections.
>
{% embed url="" %}
***
## Running a search from a taxonomy category
Once a category contains images, the Analyst can seed a new search directly from it. Instead of starting from a single reference image, the search uses all images in the category as the reference set, producing more targeted results across the full dataset.
1. From the **New Search** panel, select the **Categories** tab under **Reference Image**.
2. Select the category to search from.
3. Configure the temporal and spatial extent as needed.
4. Select **Run Search**.
A new context map appears seeded from the category. As new results are reviewed and added back into the category, the taxonomy continues to improve — analyst judgment continuously refines search precision and builds mission-specific knowledge over time.
> **Analyst** The taxonomy is a shared, living resource. Every decision adds to it, making every future search more precise. Treat it as a long-term intelligence asset, not just a way to organize a single search session.
***
## What's next
With results organized into taxonomy categories, the Analyst can review, approve, and refine the contents of each category. Continue to \[Managing your taxonomy →] to learn how to curate and build on what has been captured.
**Compliance Note** This documentation does not constitute legal or regulatory guidance. Organizations operating in classified or regulated environments should consult their compliance and security teams before deploying or using RAIC Fed. Specific data handling, classification, and audit requirements will vary by environment and mission.