Overview
This guide is for users who are accessing a DORA Prediction Map through a View-Only Share link for the first time. It outlines how to explore the map, understand the key information, and navigate the interface without needing editing access.
Jump ahead to:
What You Can Do With View-Only Access
Navigate and explore the 3D Prediction Map
View the experiment summary and key settings
Adjust the VPS display using the legend slider
Toggle visual layers on and off
Modify vertical exaggeration to enhance surface features
Expand a section by clicking below.
Interpreting the Map
Interpreting the Map
Note: The default view of the Prediction Map will vary from map to map. For quick tips on navigating DORA's interface, check out the Navigation section of this article.
The VRIFY Prospectivity Score (VPS) is visualized in the view panel, providing a clear, intuitive heatmap of potential mineralization across the area of interest. Higher prospectivity scores indicate zones where the AI model has identified strong patterns consistent with known mineral systems, helping you quickly zero in on the most promising exploration targets. This allows you to prioritize areas with the highest potential and make more informed, data-backed decisions.
Example of the visualization of VPS.
Your view-only link may have defaulted to showing target clusters as well, indicated by colour-coded groupings of VPS. We aim to cluster targets together because different data types influence predictions differently from one zone to another.
For example, targets in the north might be driven by geophysical anomalies, while those in the south could be influenced by proximity to a fault. By clustering these targets, we can more accurately analyze and evaluate the results.
Example of a visualization of target clusters.
Results will vary from one experiment to the next. In some cases, the AI will identify clear, actionable targets, visualized as VPS clusters. In other cases, the output may highlight areas of your project that warrant further exploration. Your results will vary depending on aspects like the quantity and quality of your input data, your target commodities, and the Prediction Map configurations.
In cases where data is limited, the model might detect signs of mineral potential in a zone that lies beyond known targets. Based on the available data and the AI’s feature engineering process, this zone may emerge as a high-interest area, suggesting that additional sampling or ground-truthing could be valuable.
Access and Navigation
Access and Navigation
1. Open the Link
Click the view-only share link you received. This will open the shared Prediction Map directly in the DORA interface.
Note: If you receive an access error, you may not have been granted permission. Contact the sender to request access.
2. Review a Prediction Map's Summary
When the map loads, a summary will appear automatically.
It includes:
The mineral targets and commodity thresholds
The model used in the experiment
A brief description written by the creator
You can reopen the summary at any time by clicking the notebook icon in the top-left corner of the screen.
3. Familiarize yourself with DORA's Navigation
The default view of your view-only Prediction Map link is set by the owner of the Prediction Map.
This means you may initially see VPS results, target clusters, a birds-eye-view of the satellite imagery of the area, or a similar variation. You are able to make changes to your vantage point to better inspect the map.
Use the 3D view panel to explore the map:
Left-click and drag to rotate the model
Right-click and drag to pan (move the model side-to-side)
Scroll or pinch on a trackpad to zoom in and out
Click the compass icon to reset the view to plan (top-down) orientation
Customize your View
Adjust VPS Results
Use the VPS legend slider to control the score range displayed on the map:
Toggle 3D Layers
Use the 3D Layers menu to turn individual visual layers on or off:
Adjust Vertical Exaggeration
Use the Vertical Exaggeration slider to amplify or reduce elevation differences, making topographic or subsurface features easier to see:
Understanding the Map's settings
Understanding the Map's settings
The person who configured the map will have used their industry knowledge and geoscientific expertise when configuring each setting within a Prediction Map. However, as a viewer, it can be helpful to understand the impact each step has on the overall Prediction Map process.
Expand the sections below to learn more about each step.
Step 1: Select Area of Interest
Step 1: Select Area of Interest
In this step, an Area of Interest (AOI) is set for the experiment. This step lets you set the boundary for your input features, ensuring predictions are focused within your asset or project's area.
Step 2: Input Features
Step 2: Input Features
This step is where input data is selected and added to the experiment. Your geoscientists will know which data is most relevant to the target commodity.
The data available in this step is the result of the data compilation and feature engineering process conducted by the VRIFY team using your proprietary data and available regional public data.
Step 3: Set Up Learning Data
Step 3: Set Up Learning Data
This step is where the targeted elements are defined, and where the learning data is separated into positive and negative examples, for our 2-class classification algorithm. The data files used in this step are known results such as drill assays or surface rock samples.
In this setting, the threshold indicates the minimum grade the target element(s) must meet to be represented as a positive learning data point. Element(s) that do not reach the defined grade threshold will be represented as a negative learning point.
Step 4: Embed Input Features
Step 4: Embed Input Features
In this step, your geoscientists select the most suitable AI algorithm for the project. VRIFY offers more than 10 Vision Transformer (ViT) models, each trained for specific commodities and adapted to different geological settings. A versatile Master Model is also available for use across any mineral system.
Step 5: Build Predictive Model
Step 5: Build Predictive Model
In this section, a random forest classifier is applied to your embedded data (generated from the previous step), generating your VPS results.
The configurations in this section involve telling the system how you want your data to be clustered for learning/testing purposes, which may vary depending on how varied your datasets are.
In other words, here your team controls the training and test groups that are used for validation of the model.
Step 6: Identify Targets
Step 6: Identify Targets
This final step is where your exploration targets were generated. The key output in this step is your target clusters, which identify areas of high prospectivity.
Your team typically will have started by reviewing the results from the predictive model (including quality control graphs), adjusting the parameters as needed, then review the results again. They may have repeated this cycle numerous times to refine the targets seen in this step of the map you are viewing.
From this step, view panel will display a visualization of the generated prospectivity score through a range of shades which represent the VRIFY Prospectivity Scores (VPS). This helps to refine your targets prior to exporting your final results.
Tips
Start with the Summary: It offers important context, such as the purpose of the experiment and decisions made during setup.
Use Layers Strategically: Turning off unnecessary layers can help focus on specific features.
Explore Freely: You won’t be able to modify the experiment, so feel free to zoom, rotate, and interact with the map to better understand the results.
Still Have Questions?
If you have questions or would like assistance interpreting the map, please contact:
The owner of the Prediction Map.
Your dedicated DORA representative.
Email: Support@VRIFY.com