# Harvesting Claims Using an entry or user action, you can asynchronously create _Claims_ from link annotations on steady state documents, allowing you to quickly and easily maintain your _Text Asset_ library. Vault harvests the full text, reference links, document metadata, and annotation IDs from text link and area link annotations. The harvested document metadata is based on the document fields that have the same public keys as the fields from the _Claim_ (`claims__v`) object type. To prevent creating duplicate _Claim_ records, Vault checks existing records in the _Text Asset_ library using the _Match Text_ field on the _Text Asset_ (`annotation_keywords__sys`) object. ## Configuring Claims Harvesting {#configuring-claims-harvesting} You can add the **Harvest Claims** [entry action](/en/lr/12399/) and [user action](/en/lr/12339/) to any document lifecycle state. We recommend adding the entry action to the steady state. To add the entry action: 1. Navigate to **Admin > Configuration > Document Lifecycles > \[Lifecycle] > States > \[State]**. 2. Under _Entry Actions_, click **Edit**. 3. Click **Create Entry Action**. 4. Select **Always** and **Harvest Claims**. 5. Click **Save**. To add the user action: 1. Navigate to **Admin > Configuration > Document Lifecycles > \[Lifecycle] > States > \[State]**. 2. Under _User Actions_, click **Edit**. 3. Click **Create User Action**. 4. Select **Always** and **Harvest Claims**. 5. Click **Save**. Ensure the _Claim_ (`claims__v`) object type is active. You must [activate the following object fields](/en/lr/15057/): * **Harvested Annotation ID** (`harvested_annotation_id__v`) * **Harvested Claim** (`harvested_claim__v`) * **Harvested Long Text** (`harvested_long_text__v`) * **Harvested Source Document** (`harvested_source_document__v`) You must also add the **Pending** state to the review and approval workflow for _Text Assets_. Then, add the **Send for Approval** user action to the _Pending_ state. ### Qualifying Harvested Annotations An annotation is qualified for harvest when: * The annotation is a document text link annotation or a document area link annotation. * The annotation is only linked to a reference document or anchor, as indicated by the document type being [_Suggest Links_ target-enabled](/en/lr/618/#promomats). * The annotation has more than five text characters. * The annotation text is not a duplicate of another qualified annotation's text on the same document (not case sensitive). * Text similar to the annotation does not already exist in the _Text Asset Library_, based on the volume of similar keywords found between the annotation and _Text Asset_. For example, if the annotation is "Cholecap is good" and the existing _Text Asset_ is "Cholecap is very good", it will not be harvested. ### Qualifying Harvested Claims The **Enable Claims Harvest Auto-Approve Qualified Annotations** setting is enabled by default. When enabled, Vault automatically places qualified harvested _Claims_ in the _Approved_ state. A harvested _Claim_ is qualified automatically for approval when: * The source document is in its steady state, such as _Approved for Distribution_. * The linked document is in its steady state, such as _Approved for Use_. * The source annotation is a text annotation. * The annotation text is less than 1,500 characters. To disable this functionality, navigate to **Admin > Settings > Application Settings** and, under _Text Asset Library Management_, clear the **Enable Claims Harvest Auto-Approve Qualified Annotations** checkbox.
## Limitations {#limitations} The following limitations may affect _Claim_ harvesting: * Vault does not harvest _Claims_ if the document is a binder. * Vault does not harvest _Claims_ from video or audio files. * Vault does not harvest _Claims_ from ZIP files. * Vault may fail to harvest _Claims_ from area annotations if the selected area is small or lacks detectable text. If the text cannot be selected while in [Annotate mode](/en/lr/9777/#Annotate_Mode), it is unlikely that text is detected. A qualified annotation will not become a harvested _Claim_ if any of the following items are true: * The document metadata does not provide a value for _Country_ (`country__v`) and _Product_ (`product__v`). * The _Claim_ (`claim__v`) object type of the _Text Asset_ (`annotation_keywords__sys`) object has required fields designated by the customer for a record to be created, and the document metadata does not provide a value for the required fields. * The document type must have document fields with the same public key as the required object record fields, and the data type of the fields is the same. ## Related Permissions The following permissions control functionality related to harvesting _Claims_: | Type | Permission Label | Controls | | --- | --- | --- | | Security Profile | Objects: Text Asset: Create, Edit | Ability to harvest a _Claim_ from a document| | Security Profile | Objects: Text Asset: Object Field Permissions: Primary Country (`country__v`), Primary Product (`product__v`): Read | Ability to harvest a _Claim_ from a document | | Security Profile | Objects: Text Asset Brand: Create, Edit | Ability to harvest the additional brands from a document as _Text Asset Brand_ child records of the harvested _Claim_ | | Security Profile | Objects: Text Asset Country: Create, Edit | Ability to harvest the additional countries from a document as _Text Asset Country_ child records of the harvested _Claim_ | | Security Profile | Objects: Text Asset Indication: Create, Edit | Ability to harvest the additional indications from a document as _Text Asset Indication_ child records of the harvested _Claim_ |