After you configure the required data sources, you can use the product's built-in models to create datasets for your enterprise, such as user tag and behavioral datasets. These datasets provide the foundation for subsequent audience selection and marketing campaigns.
You can create the following types of datasets:
User tag dataset: Build a user tag dataset based on a large wide table of user tags. You can then use this dataset for analysis and audience tag filtering. You can also push the tags to Brand Databank.
You can upload tag table files in XLS or XLSX format. This lets you use your offline user tag data.
RFM Dataset: Build a Recency, Frequency, and Monetary (RFM) model based on transaction or customer consumption data. This model analyzes consumer purchasing power across three dimensions: recency, frequency, and monetary value. You can then use this model for RFM analysis and audience RFM model filtering.
AIPL Dataset: Build an AIPL model based on a behavioral dataset or customer statistics. This model segments your brand's audience and quantifies your audience assets. It is a critical part of your brand's omni-channel marketing. After you build the model, you can use it for AIPL analysis, AIPL flow analysis, and audience AIPL model filtering. You can also push the AIPL model to Brand Databank.
Behavioral dataset: Build a behavioral dataset based on user behavior data. You can use this dataset for behavior filtering and AIPL model building, and for repurchase prediction and product recommendation.
Product tag dataset: Build a product tag dataset based on a large wide table of product tags. You can use this dataset for product recommendation.
The following figure shows the data links for each type of dataset.
Create a dataset folder
You can create dataset folders to organize your datasets for easier management. You can also create multiple levels of subfolders.
Procedure:
Go to Workspace > Configuration Management > Data Center > Dataset to open the dataset management page.
On the dataset management page, click New > Folder in the upper-right corner.
To create a subfolder, first click a folder name in the list to open it. Then, click New > Folder.
In the dialog box that appears, enter a folder name and click Confirm.
To view the datasets stored in a folder, click the folder name in the list.
You can search for datasets by keyword, regardless of whether they are in a folder.
You can rename or delete folders:
Rename: Click the
icon. In the dialog box that appears, enter a new name and click Confirm.Delete: You can only delete empty folders. Click the
icon and confirm the deletion.