After you report user data, item data, and behavioral data, you can use the data management feature of the console.
1. You can check whether data is uploaded to AIRec. If the value true is in the response after you use server SDKs to push data, the data is uploaded. Then, you can view the data details on the Update Message Query page. For more information, see the "Update Message Query page" section.
2. If you want to view the information about a user or item entry, you can specify fields on the Data Query page for a query. For more information, see the "Data Query page" section.
3. You can view the metrics that are related to the quality of data on the Data Quality Report page 1 hour after data is uploaded. For more information, see the "Data Quality Report page" section.
1. Data Quality Report page
After you upload data, you can check whether exceptions occur on the data. If an exception occurs, click Fix Now to view and handle the exception. Note: To improve service stability and prevent performance deterioration, we recommend that you handle an exception on the Data Quality Report page immediately after the exception occurs. You can also view the total number of behavior entries on the Data Quality Report page. To view the details about the behavior entries, see the "Update Message Query page" section. A data quality report contains metrics such as the number of items, repetition rate, rate of valid data, data integrity, and user logon rate. A statistical report of full data is also available to display derived metrics such as the number of exposures, number of clicks, CTR, UV, and PV. You can check whether your data quality and business metrics meet your business requirements based on the measurement results of these metrics. If they do not meet the requirements, identify issues as required. The data quality report displays data information about 1 hour after baseline or incremental data is uploaded.
Number of items
Description: the number of raw items in a table. Duplicate records are counted. However, the items that are used are the items after deduplication. To view the details, go to the Basic Information page in the console.
Usage: This metric helps you check the historical data that you uploaded and compare the value of this metric with the quota that you purchased.
Description: the proportion of duplicate data entries in a table. The following table describes the primary keys that are used to identify whether a data entry is a duplicate data entry. A unique identifier is a combination of fields.
user_id and user_id_type
item_id and item_type
Usage: The data that is used is the data after deduplication. A high repetition rate leads to a small amount of valid data.
2. Data Distribution page
After data is uploaded, you can view the total number of behavior entries within the last three days, the total number of users, and the total number of items in the upper part of the Data Distribution page. You can also view the metrics that are related to the distribution of item data in the Type-based Item Distribution section of the Data Distribution page.
Recommendable Items by Scene ID
Recommendable Items by Type
Recommendable Item by Type for Each Scene ID
Proportion of Recommendable Weighted Items for Each Scene ID
Proportion of New Items for Each Scene ID
Proportion of Temporary Items for Each Scene ID
3. Data Query page
You can view specified user data or item data on the Data Query page. To check whether an item is uploaded or to view an uploaded item, you can click the Item Data tab on the Data Query page, specify Item ID and Item Type, and then click Search to query the details. To check whether a user entry is uploaded or to view an uploaded user entry, you can click the User Data tab on the Data Query page, specify User ID, and then click Search to query the details.
4. Update Message Query page
After you upload behavioral data, user data, and item data, you can view the upload records of these types of data on the Update Message Query page.