Custom Attributes

Prev Next

Available in VPC

Custom Attributes uses NCLUE features to infer similarity between user attributes that you define directly as strings.
You can use the Custom Attributes menu interface to create and delete Custom Attributes Profiles.

Custom Attributes interface

The Custom Attributes interface includes the following components:

nclue-custom-vpc_screen_ko

Component Description
① Menu name Current menu name.
② Basic features Basic features provided in the Custom Attributes menu.
  • [Create profile]: Click to create a profile
  • [Features and pricing information]: Click to go to the NCLUE service overview page.
  • [Refresh]: Click to refresh the profile list.
Delete Click to delete a profile
④ Profile list List of created profiles
  • Profile ID: Unique ID automatically assigned when the profile is created.
  • Profile name: Name entered when the profile was created.
  • Status: Status of the profile.
    • Registered: The profile creation process has started.
    • Pending: Waiting before proceeding to the next step. If there are many requests, the wait time may be longer.
    • In progress: The profile is currently being created.
    • Available: The profile has been created and is now available.
    • Failed: Profile creation failed.
  • Result data bucket: Name of the Object Storage bucket where the profiled result data is stored.
  • Result data file path: Path to the file storing the profiled result data.
  • System message: Message shown when creation fails (see NCLUE Troubleshooting).
  • Creation date: Date and time when profile creation was requested.
  • Completed at: Date and time when profile creation was completed.
  • NCLUE model: Name of the NCLUE model used for feature creation.
  • Feature: Name of the feature used for profile creation.
  • Row count: Number of rows contained in the feature used by the Task.
  • Attribute keywords: Keywords used to check similarity between users contained in the feature.

Create a profile

Create a profile that identifies similarity between user Features and attribute keyword Features. To create:

  1. In the VPC environment of the NAVER Cloud Platform console, navigate to Menu > Services > AI Services > NCLUE.
  2. Click the Custom Attributes menu.
  3. Click [Create profile].
  4. When the create profile popup appears, configure the profile.
    • Profile name: Enter a unique name within 100 characters.
    • Feature: Select the feature to be profiled from the dropdown menu.
    • Attribute keywords: Enter the keywords for which you want to measure similarity with target users.
      • Enter keywords or phrases that users might naturally search on NAVER.
        • Example: If you want to identify attributes of public transportation users:
          • <Keywords (separated by line breaks)>
            K-Pass
            Public transportation
      • You may enter up to 1,000 characters in total; separate each keyword with a line break
      Note
      • All entered attribute keywords follow an AND search rule. You cannot set AND/OR conditions.
      • It is recommended to enter keywords that clearly represent the characteristics of the target users.
      • To find users who do not have certain characteristics, enter that keyword and look for users with low similarity.
      • If the target user's keywords contain different traits, it is recommended to enter and profile them separately rather than all at once.
      • There are no internal indicators or recommendations for keyword selection.
    • Result storage path: Result storage path
      • Storage bucket: Choose the Object Storage bucket from the dropdown menu.
        • To use a new Object Storage bucket, click [Create an Object Storage]. After creating i-nclue_refresh, click the refresh icon (i-nclue_refresh) so Object Storage appears in the dropdown. For details, see Object Storage Bucket Creation.
      • File path: Enter the path where the Object Storage result file will be stored in the bucket
        • The following special characters cannot be used in file paths:
          • &$@;:+,?*\{}^%`[]<>~#|"'
          • Characters with ASCII codes 0-31 and 128-255 are also not allowed.
          • Spaces
        • File naming rules:
          • If the input string ends with /, it is treated as a directory. A file named nclue_{random}_{YYYYMMDDHH}.tsv will be created inside the directory.
          • Otherwise, the input string is treated as a filename prefix, and a file named {input}_{random}_{YYYYMMDDHH}.tsv will be created.
  5. Click [Create].
    • The profile appears in the list with the status Registered. When profile creation is complete, the status changes to Available.
Note

Profile results are produced as values between -1 and 1. A value closer to -1 indicates lower similarity, and a value closer to 1 indicates higher similarity.

Caution
  • Result data is saved as a TSV file in the storage path specified when the profile was created, not in the service's internal storage. Object Storage usage fees will apply depending on the file size.
  • Check the version of the NCLUE model. Profiles created from features based on deprecated models can be viewed if they have been downloaded to the result bucket, but they may differ from results generated with the latest models. We recommend using the latest models.

Delete a profile

To delete a profile:

Caution
  • Deleted profiles cannot be restored. Be sure to review the profile before deleting it.
  • Even if you delete a profile, the result data stored in Object Storage is not deleted, and Object Storage fees will continue to occur. Delete the Object Storage files separately to avoid charges.
Note
  • If you select a profile whose status is Pending or In progress, [Delete] is disabled.
  • You can select and delete multiple profiles at once.
  1. In the VPC environment of the NAVER Cloud Platform console, navigate to Menu > Services > AI Services > NCLUE.
  2. Click the Custom Attributes menu.
  3. In the profile list, select the profile you want to delete, then click [Delete].
  4. When the Delete profile popup appears, click [Delete].
    • The profile is deleted and removed from the list.