> For the complete documentation index, see [llms.txt](https://alphamoon.gitbook.io/documentation/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://alphamoon.gitbook.io/documentation/user-guide-how-tos/data-points-definition.md). # Data points definition You can **tailor the data extraction** to your needs by editing already existing data points or creating new ones. Want to know how to do that? Let's dive in! ### Start with the Configuration First, select one of your processes from your Dashboard. Once you are in the **Queue** view, click the [Configuration](/documentation/platform-walkthrough/configuration.md) button:

In the Configuration, from the navigation on the left side of the page, select the **Data points definition** tab**.** {% hint style="info" %} **Note:** When you create a custom process, you will be redirected to the **Data points definition** automatically. {% endhint %} ### Data points used in the process For the supervised processes (Invoices, Receipts, Passports) and templates (e.g. Driving licenses, Purchase orders, Birth certificates, Utility bills), we have already set a ready-made field selection for extraction. Take a look:

{% hint style="info" %} **Custom process** A part of **creating your custom process** is setting up your custom fields from scratch, therefore the data points list will be empty. {% endhint %} ### Add new data points You can also add new data points to your process. To do that, click the **+Add new data point** button.

### Supervised processes In the case of supervised processes (Invoices, Receipts, and Passports), you can decide whether you wish to capture a **Custom** or **Default** data point. * Choose the **Default** option whenever you want to add ready-made fields for extraction from the dedicated list. * Choose the **Custom** option whenever you want to define a new data point from scratch. \ **Default data point**

* The first step is to select the type of data you want to capture. In the case of the supervised processes, you can capture both text and [tables](/documentation/user-guide-how-tos/table-data-extraction.md) (the columns, data types, and relationships between columns, as well as the actual information stored in the table). * Next, choose the ready-made field from the list. * Now you can either hit the **Create new field** button on top of the panel or go through the **Optional settings**:
* **Data format**\ Select what formatting should we apply to this data point when found in a document. [Learn more about converters](/documentation/user-guide-how-tos/converters.md).
* **Occurrence**\ If the field is unique for your document, select 'One'. If it appears more than once (e.g. you want to capture 'Date' and there are several dates within the document), go with the 'Many'.
* **Mandatory**\ Select ‘No’ if capturing your new field is not necessary for every document to have an **Accepted** status. Go with ‘Yes’ in case you wish to see the **Needs review** status every time your field won't be captured from your documents automatically.
* **Confidence level**\ A value of 0.9 means that when our platform captures your field with a confidence less than 0.9, your file will get the **Needs review** status. You can enter the confidence value between 0 and 1. #### **Custom data point**

* The first step is to select the type of data you want to capture - text or table. * Next, describe the field you want to capture from the document, e.g., Date, Name, or Total amount. We will use this label to automatically pull the described data from your files. * Decide whether you want Alphamoon to capture data automatically. Select 'Yes' to automatically capture this field from your documents. If you choose 'No', the model will not pull this field automatically. However, you can train the model to capture this information from your documents. Turn on continual learning in the configuration and mark this field correctly in several documents that you upload. * Now you can either hit the **Create new field** button on top of the panel or go through the **Optional settings** ([described above](#supervised-processes)). ### Templates In the case of templates (e.g. Purchase orders, Driving licenses, Utility bills), you can add your **Custom data points** to the ready-made field selection.

* In the templates, there is only one type of data - text. * Define data point name - describe the field you want to capture from the document, e.g., Date, Name, or Total amount. We will use this label to automatically pull the described data from your files. * Now you can either hit the **Create new field** button on top of the panel or go through the **Optional settings** ([described above](#supervised-processes)). ### Custom process After you create the Custom process, we will automatically redirect you to the **Data point definition**. [Learn more about creating a custom process](/documentation/user-guide-how-tos/processes/how-to-create-a-custom-process.md).

* In the custom process, there is only one type of data - text. * Define data point name - describe the field you want to capture from the document, e.g., Date, Name, or Total amount. We will use this label to automatically pull the described data from your files. * Now you can either hit the **Create new field** button on top of the panel or go through the **Optional settings** ([described above](#supervised-processes)).
{% hint style="warning" %} **Changes to Settings**\ Bear in mind that changes to the data points settings will only apply to files sent from the moment these changes are made. {% endhint %} {% hint style="info" %} **Need support?** If you have any questions regarding the data points definition, feel free to contact us at: **.** {% endhint %} #### #### See next: