Available in VPC
With NCLUE of NAVER Cloud Platform, anticipate user behavior, gain a better understanding of their attributes, and apply the insight into marketing. You can see how to use it in detail in Getting Started with NCLUE and Using NCLUE, but we recommend that you first go through the full quickstart document. Reading the guide after learning the quickstart document will allow you to use NCLUE more efficiently. Check a quickstart document that best suits your purpose of using NCLUE.
- Task Model execution quickstart: when you have comprehensive data and a target dataset, create and execute Task (Task Model, machine learning model)
- User profiling quickstart: if you want to identify user attributes previously unknown to you based on your sequence dataset, create Profile to see users' shopping interests and similarities by keyword
Task Model execution quickstart
In case you want to execute the Task Model, the overall sequence of using NCLUE and the description of each sequence are as follows:
1. Preparing dataset
Prepare the sequence dataset and the target dataset required to run the Task Model and store them in the Object Storage bucket. If you prepare multiple target datasets depending on your purposes, various Task Models can be created with the same sequence dataset. You can refer to the following user guide:
2. Creating Feature
Create Feature using the prepared sequence dataset. Feature is a fixed vector value converted from a user's behavior history. It can be used to create Tasks or Runs. You can refer to the following user guide:
3. Creating Task
Create Task based on the target dataset and the sequence dataset that was used to create Features. You can refer to the following user guide:
4. Creating Run
Create Run to execute the Task Model. You can apply a new sequence dataset to the Task Model created in Step 3. You can refer to the following user guide:
User profiling quickstart
Create Profile of Shopping Intent or Custom Attributes depending on your purpose of use. If you create Shopping Intent Profile, you can see the information of NAVER users that are similar to the Task Model's target group such as gender, age, and presumed shopping interests. If you create Custom Attributes Profile, you can have keywords that best describe the target user and check the similarities.
Shopping Intent quickstart
In case you want to create Shopping Intent Profile, the overall sequence of using NCLUE and the description of each sequence are as follows:
1. Preparing dataset
Prepare the sequence dataset and the target dataset required to profile a user and store them in the Object Storage bucket. If you prepare multiple target datasets depending on your purposes, various Task Models can be created with the same sequence dataset. You can refer to the following user guide:
2. Creating Feature
Create Feature using the prepared sequence dataset. Feature is a fixed vector value converted from a user's behavior history. It can be used to create Tasks, Shopping Intent Profiles, or Custom Attributes Profiles. You can refer to the following user guide:
3. Creating Task
Create Task based on the target dataset and the sequence dataset that was used to create Features. You can refer to the following user guide:
4. Creating Profile
Select the Task Model to create Shopping Intent Profile. You can refer to the following user guide:
Custom Attributes quickstart
In case you want to create Custom Attributes Profile, the overall sequence of using NCLUE and the description of each sequence are as follows:
1. Preparing dataset
Prepare the sequence dataset required to profile a user and store it in the Object Storage bucket. You can refer to the following user guide:
2. Creating Feature
Create Feature using the prepared sequence dataset. Feature is a fixed vector value converted from a user's behavior history. It can be used to create Tasks, Shopping Intent Profiles, or Custom Attributes Profiles. You can refer to the following user guide:
3. Creating Profile
Input keywords that you want to infer the similarity with the user's Feature to create the Custom Attributes Profile. You can refer to the following user guide: