Attributes Basic Overview

About Attributes Basic Overview

An attribute is a property of a document that characterizes it in some way. Common examples of attributes include the author name and creation date.

You can create custom attributes for your projects. There are also a variety of system attributes for you to use. You can also set up intelligent entities which automatically detect attributes based on the document’s text (for example, when a brand or a product is mentioned in a customer’s feedback).

After you add attributes, it’s possible to create additional derived attributes to further understand your data. You can also organize your attributes into attribute sets, making it easier to report on these attributes.

Attribute Field Types

The following field types are supported for attributes:

• Text
• Number
• Date

Accessing Attributes

Attributes are managed on a project level. To access the attributes for a project:

1. In Designer, go to the Admin tab and find the project you’re interested in.
2. Click the Attributes option in the Actions column.

This opens the attributes table containing the following information about each attribute in the project:

• Name: The attribute’s system name. See Attribute Naming for naming requirements.
• Display name: The attribute’s display name, which appears in reports, filters, etc. See Attribute Naming for naming requirements.
• Type: The attribute’s type. For standard attributes, values can be Text, Number, or Date. For derived attributes, values can be Dimensional Lookup, Range Rollup, Satisfaction Score, or Derived From Category.
• Group: The attribute’s group that represents its origin and intended use. Values be one of the following:
  • Category derived: attributes derived from models or categories.
  • System: system attributes.
  • Customer-defined: all custom attributes available from the data source of your choice (including dimensional lookups, range roll-ups, and satisfaction scores).
  • Scorecard: attributes used in intelligent scoring.
• Enable for Reporting: Shows if an attribute is enabled for reporting (On) or not (Off).
• Case Sensitive: Shows if an attribute is marked as case-sensitive when displaying values in Document Explorer, Source Highlighter, Custom Export, and sentences preview export.
• Drill Path: Shows custom drill path if it was defined. If there is no custom drill path, then this field will be blank.
• E-mail field: Shows if an attribute contains an email address.
• Enable DocValue: Shows if ElasticSearch doc values are used for this attribute.
• Enable delimited multivalue: Shows if multiple values are enabled for this attribute.
• Parent: If the attribute is a derived attribute, then this field will display “parent.” This field will be blank for custom and standard attributes.
• Actions: Perform the following actions on the attribute:
  • Edit the attribute
  • Create a derived attribute
  • Delete the attribute

Managing Attribute Sets

Use the Attribute Sets toggle at the top of the page to view your attribute sets. This allows you to create new attribute sets and delete existing ones. Select Attributes to view your individual attributes.

System Attributes

There are a number of system attributes, such as Document Date and Source ID, that are automatically applied to each document uploaded to XM Discover. These attributes help you manage feedback within XM Discover, as well as enrich it with XM data derived by the NLP engine.

Below is a table of the different system attributes, grouped by the different categories of attributes. This table contains the following information about each attribute:

• Name: The attribute name that appears in reports, filters, etc.
• System Name: The attribute’s system name which you use to query or filter your data.
• Type: The attribute type.
• Description: A brief description of the attribute’s meaning and purpose.
• Granularity: The level of data granularity associated with an attribute. For example, Sentence Word Count is only relevant on a sentence level, while Document Date is available both for a document and for each sentence in that document.

IDs and References

Date and Time

Word Counts and Position