Label Management

---

Overview

LightBeam Spectra's Label Management feature enables users to classify and organize documents using predefined labels. Labels are tags with a priority and definition that can be applied to documents to categorize them based on their content or sensitivity. Labels can be created natively within LightBeam or integrated from third-party systems like Microsoft Information Protection (MIP).

By using labels, users can easily manage access, distribution, and retention of documents based on their classification. The definition of labels depends on each company's policies.

LightBeam automates the labeling process by:

1. Scanning documents

2. Identifying sensitive attributes (e.g., SSNs)

3. Analyzing file classifications

4. Automatically applying the appropriate label

---

Key Terminology

• Label: A tag with a priority and definition that is applied to a document to classify it based on content or sensitivity.

• Label Set: A collection of labels that can be used together in a policy. Label sets can be of multiple types, including LightBeam label sets and MIP label sets.

• Labeling Policy: A set of rules and conditions that determine which labels should be automatically applied to documents.

• Object: A file, document, spreadsheet, etc. in a data source.

• File Classification: The categorization of a document based on its content and context, as determined by LightBeam Spectra's machine learning algorithms.

---

Labels & Label Sets

Labels are organized into Label Sets within LightBeam. A Label Set is a collection of labels that can be used together in a policy. Each label within a Label Set has the following properties:

• Name: A descriptive name for the label.

• Priority: A numeric value that determines the order in which labels are evaluated and applied. Higher priority labels take precedence over lower priority ones.

• Definition: The conditions that a document must meet for the label to be applied. Definitions can include attributes like keywords, regular expressions, data types (e.g., SSN, name, etc.) or file classifications.

Label definitions support the following operators:

• Any of: The document must contain any one of the specified attributes or belong to any of the specified file classifications.

• All of: The document must contain all of the specified attributes and belong to all of the specified file classifications.

• None of: The document must not contain any of the specified attributes and must not belong to any of the specified file classifications.

• Not all of: The document must not contain all of the specified attributes and must not belong to all of the specified file classifications.

Label definitions can also include multiple conditions combined using the "AND" or "OR" operators. The same operator must be used between all conditions in the definition.

---

Creating a Label Set

To create a new Label Set:

1. Click on the Insights option in the header menu.

Figure 1: Click on Insights

1. Select the "Label Management" icon from the left sidebar menu.

Figure 2: Label Management icon

1. Click on the "Create LightBeam Set" button in the top right corner of the page.

Figure 3: Label Management page - Create LightBeam Set button

1. In the Create LightBeam Label dialog box, enter a name for the Label Set.

Figure 4: Create a LightBeam Label Set

1. While creating a new Label Set, click on the "Add Label" button to define labels directly.

   If you have already saved the Label Set without adding labels, you can click on the "Create a new label" button within the saved Label Set's details page.

Figure 5: Create a new label within new set

• Enter a name in the "Label Name" field.

Figure 5.1: Define label properties

• Specify the priority using the "Priority" dropdown menu.

• Define the "Label Conditions" within the given section by selecting attributes, operators, and values.

1. Once all the labels are added, the Label Set will appear as shown in Figure 5.2

Figure 5.2: Label Set with added labels

---

File Classification-based Labels

Alerts 2.1.1 introduces the capability to define labels based on file classifications in LightBeam Spectra. File classifications are determined by LightBeam Spectra's machine learning algorithms, which categorize documents based on their content and context.

1. Click on the Insights option in the header menu.

Figure 1: Click on Insights

1. Select the "Label Management" icon from the left sidebar menu.

Figure 2: Label Management icon

1. Click on the "Create LightBeam Set" button in the top right corner of the page.

Figure 3: Label Management page - Create LightBeam Set button

1. In the "Create LightBeam Label" dialog box, write "Classification Labels" for the Label Set name.

2. To create file classification-based labels within this label set, click on the "Create a new label" button within the saved Label Set's details page.

3. Enter a name in the "Label Name" field, for instance "All Financials", "Tax Forms", etc.

4. Specify the priority for the label using the "Priority" dropdown menu. In this example, since the "Tax Forms" label has a higher priority than "All Financials," we will set its priority as "2" and set "All Financials" to "1."

   Once all the labels are added, the Label Set will appear as shown in Figure x.

   

1. To define the Label Conditions, click on "View Details" corresponding to the Label Name under the "View Definition" column. This will display a window where you can specify the file classification conditions for the label.

2. This will display a window as seen in Figure. Under the "CONDITIONS" subsection choose "File Classification" from the drop-down menu as the condition type for defining the label conditions.

• After selecting "File Classification," you will see options to specify the operator and value for the condition.

• Choose the appropriate operator from the dropdown menu, such as "Any of these (OR)," or "Not all of these(NOR)". This operator determines how the file classification conditions will be evaluated.

1. Click on Select File Classification.

1. In the file classification selection dialog, you will see the available file classifications.

   • Select the desired file classifications by checking the corresponding checkboxes, like "Financial". You can choose multiple classifications from different categories if needed.

   • On selecting, the relevant categories will expand to display subcategories from which you can again choose multiple file types, such as "Tax Forms".

   • Then, click on the Add button.

   
2. You can add more attribute-based conditions by clicking on the "Add more conditions". Note that all file classification conditions must be added in the previous selection dialog, as multiple file classification conditions cannot be added separately.

1. Once you have finished defining all the necessary conditions, review the label details. If everything looks correct, click on the "Save" button to create the file classification-based label with the specified conditions.

1. Now repeat the same process for other label sets.

Verifying File Classification-based Labeling

By comparing the counts in the Classification section with the counts on the label cards, users can verify that the file classification-based labeling is functioning as expected.

Step 1: Review the Classification Counts

To verify the application of file classification-based labels to documents, follow these steps:

1. Go to Insights in the LightBeam Spectra console

2. Navigate to the "Classification" option in the left sidebar menu

1. The "Classified Documents" section provides an overview of the file classifications and their respective counts. Focusing on the "Financial" classification, you can see several sub-classes, including "Finance Others," "SEC," "Earning Statement," "Invoices / Receipts," and "Tax Forms." These sub-classes represent more specific categories within the "Financial" classification.

2. Take note of the count for the "Tax Forms" sub-class, which is 373 in the provided example.

3. Additionally, observe the counts for the other financial-related sub-classes, such as "Finance Others," "SEC," "Earning Statement," "Invoices / Receipts," etc.

Step 2: Verify the Label Card Counts

1. Click on the Insights option in the header.

2. In the left sidebar menu, locate and click on the "Label Management" option.

Figure 2: Label Management icon

1. Within the Label Management interface, click on the "Label" tab to view Label cards.

Here, you will be able to view all the File Classification-based labels cards.

1. In the "Labels" section, locate the label cards representing the file classification-based labels

2. Identify the "Tax forms" label card and check its count. In the example, the count is 373.

3. Next, locate the "All Financials" label card and check its count. In the example, the count is 3058

Step 3: Compare the Classification Counts with the Label Card Counts

1. Verify that the count for the "Tax Forms" sub-class in the Classification section matches the count on the "Tax forms" label card. This indicates that all documents classified as "Tax Forms" have been correctly labeled with the corresponding "Tax forms" label.

2. Ensure that the count for the "All Financials" label card is equal to the sum of the counts of all the other financial-related sub-classes in the Classification section (Image 2), excluding the "Tax Forms" sub-class. This ensures that the "All Financials" label accurately represents the collective count of all other financial documents.

Note: Please note the following constraints and limitations when using file classification in label definitions:

• File classification conditions can only use the "OR" and "NOR" operators, as a document can only have one file classification assigned to it.

• File classification can only be specified once in the label definition, as multiple classifications cannot exist together on a single document.

File classification-based labels can be used in Labeling Policies to automatically apply labels to documents based on their file classification. Refer to the File Classification-based Policie section in the Policy and Alerts document for more information on creating and applying these policies.

---

Label Application

Manual Label Application

Users can manually apply labels to documents within LightBeam. This allows for flexibility in cases where automated labeling may not be sufficient or when users need to override the assigned labels.

Users can manually apply labels to documents within LightBeam:

1. Navigate to the Object Viewer page for a specific document (Figure 6).

Figure 6: Object viewer page

1. On the object viewer page, locate the "Labels" section.

2. Click on the "Edit Labels" button within the "Labels" section to open the "Edit Labels" dialog.

Figure 7: Labels section on object viewer page

1. In the "Edit Labels" dialog, click on the arrow next to the desired label set for a drop-down menu.

Figure 8: Edit Labels dialog

5. Select the required label from the available options by checking the box next to the label name.

Figure 8.1 : Edit Labels dialog box drop-down menu.

1. Click on the "Save" button to apply the selected labels to the document.

Manually applied labels will be visible in the document metadata and can be used for filtering and searching documents. This provides a convenient way to categorize and locate specific documents based on their assigned labels.

---

Automatic Label Application via Labeling Policies

LightBeam supports automatic label application through Labeling Policies. Labeling Policies allow administrators to define rules and conditions that determine which labels should be applied to documents automatically.

Creating a Labeling Policy

1. Click on "Playbooks" in the header menu.

Figure 9: Playbooks in header menu

1. Scroll down and locate the Labeling policy section. Click on the "Create New" button within the Labeling policy box.

Figure 10: "Create New" button within Labeling policy

1. On the next page

• Enter a name for the Labeling Policy under Rule Set Name.

Figure 11: Create New Rule Set

• Click on Select Rule Set Criteria to choose a rule set for the policy from the available Label Sets.

Figure 12: Step 1: Select Entities & Attributes

• Click on Next.

Figure 12.1: Step 1 - Select Entities & Attributes

1. Here, choose the data sources to which the policy should be applied by ticking the checkboxes next to the desired datasources.

Figure 13: Step 2 - Select Data Sources

Figure 13.1: Step 2 - Select Data Sources

Review the policy settings, including the selected Label Set and data sources.

1. Click on Save & Close to create the Labeling Policy.

When a Labeling Policy is executed, LightBeam evaluates each document against the label definitions in the specified Label Set.

---

Viewing Label Statistics and Objects

In the Label Management section, you can view statistics about the labels and the number of objects associated with each label. This information is presented in a card view format, providing a quick overview of your labeled data.

1. Navigate to the Label Management page.

2. Click on the Labels tab.

3. On the Label Management page, you will see a card view displaying all the available labels (Figure 14).

Figure 14: Label statistics and object counts

1. Each card represents a label and shows the count of objects associated with that label.

Figure 15: Linked LB and MIP Label Sets

1. If a label is linked to an MIP (Microsoft Information Protection) label set, you will see an indication that it's linked. The MIP label set name will be displayed along with the LightBeam label set name (Figure 15). For example, let's take a closer look at the "Restricted - MIP" label card:

• The "Restricted - MIP" label is linked to the "LB Linked LB set - MIP" LightBeam label set and the corresponding MIP label set.

• The card shows that there are 57 objects associated with the "Restricted - MIP" label.

• The link(🔗) icon next to the label name indicates that it is an MIP-linked label.

1. To view the objects associated with the "Restricted - MIP" label, click on the label card and then click on Objects.

Figure 16: Objects associated with a label set

1. You will be taken to a page displaying the following information for each object labeled as "Restricted - MIP":

• The data source name where the object resides

• The data source type where the object resides

• No. of objects in each data source

Figure 16: Details of objects associated with a label set

1. You can also use the filter options on the page to refine the list of objects based on specific criteria, such as the label name, data source, or date range (Figure 16.1). In the object's metadata, you will see the labels applied to that object.

Figure 16.1: Details of objects associated with a label set

1. If you click on the file link (if you have permission), you will be taken to the actual file, where you can see the MIP label applied.

---

Third-party Integration - MIP

Currently, LightBeam integrates with Microsoft Information Protection (MIP) as its primary third-party label management solution. This integration allows for seamless synchronization of label definitions and policies between MIP and our system.

System Compatibility

The system is designed to operate exclusively with LightBeam label sets. The labeling policy will evaluate the LightBeam label set, and users will be required to provide definitions for these labels. The terminology employed will be specific to the LightBeam label sets, and they will be mapped to the corresponding MIP label sets.

Label Integration

In the event that an MIP label set has been previously onboarded and associated with a LightBeam label set, users will have visibility into the labels originating from MIP as well as those linked to LightBeam. While the label names can be customized within LightBeam, the labels imported from MIP are externally derived and cannot be modified within our system. Our system's capability is limited to mapping the MIP labels to their respective LightBeam counterparts.