To enable the Lark dataset, you must first complete Lark Integration in “2. Basic configuration”.
Lark base table is a business management tool based on tables, allowing you to manage projects from multiple dimensions and supporting various summary views. Therefore, in project management and business management scenarios, you may use Lark base table instead of Lark Sheet for data recording. Currently, Lark base table supports integration with datasets for data analysis.

(1) In the dataset module under data preparation, click “Create” at the top of the dataset list on the left, then select “Create dataset”.

(2) Select to connect to “Base table”.

(3) Enter the Lark base table link and click “Confirm” to create the data connection.

(4) In the data connection, locate and click to select the Lark base table you just integrated.
(5) Enter the dataset configuration interface to configure settings.
For dataset configuration, refer to Field Configuration. After saving and completing the creation of the dataset, data synchronization will occur according to the set synchronization frequency. In the Lark base table dataset, each synchronization will retrieve all data from the corresponding model and write it into the current partition.

3.1 Supported field types #

Lark base table supports various field formats, including text, single choice/multiple choice, date, person, attachment, and more. These field formats will be mapped when integrating with the dataset. Among them, attachment, association, lookup reference, and formula fields are not supported for dataset integration.

3.2 Time zone description for date/time type fields #

Date/time fields integrated into the dataset are consistent with the timestamps of date/time fields in Lark base table.
Specifically, in Lark base table, the local time is displayed according to the user's local time zone. However, after integrating intelligent data insights to generate a dataset, the corresponding time will be displayed based on the data center where the dataset is located. By default, the time is shown as UTC+8 (original time plus 8 hours), which typically means adding 8 hours to the default original time zone of Lark multi-dimensional tables, resulting in a time format that aligns with the China time zone.

Timestamp: Also known as Unix timestamp (Unix Timestamp), it is a method of representing date and time as a numeric value.

Format: Convert time or text-type time into a timestamp using toUnixTimestamp(xxx), that is, convert yyyy-mm-dd hh:mm:ss into a timestamp.
Example:

Extension: For the converted timestamp, add the timestamp difference of 28,800 seconds corresponding to 8 hours to complete the conversion from UTC+0 to UTC+8, then use the toDateTime function to convert it into a more readable time format.

3.3 Notes #

Permissions and functionality:

• The dataset owner must have read permission for Lark multi-dimensional tables.
• After the user enters a link to a Lark table, all data tables under that table will be imported.
• Table formulas are currently not supported.
• Importing multiple sheets from ordinary Lark tables is currently not supported.
• If there are empty rows, they will be counted in the row statistics ([null,null,null,null] format data will be included in the dataset).

Naming conventions:

• When naming Lark documents or table files, please follow naming conventions and try to avoid using the English period '.' in table names and sheet names.
• Column names: When naming columns, please follow naming conventions and use plain text as much as possible. Avoid column names that are purely numeric, images, or hyperlinks. Also, avoid naming columns as p_date or including $, `(backtick), ! and other regular expressions. Since the column names captured during synchronization are those at creation, please do not modify them arbitrarily after creating the dataset. Column names must be unique.