It is not clear what kind of data these are, but assuming they are workflow processes, the following is a possible temporal profile of the given processes:

* 20% of fine creation and send processes go through an appeal to judge (frequency = 20), while 50% go through a payment process (frequency = 20). The remaining 30% go through an insertion of a fine notification, which may also include an appeal to judge or a payment process.
* An additional 10% of fine creation and send processes include a date appeal to the prefecture, which is inserted after the fine is sent and before any payment or appeal to judge (frequency = 20).
* An additional 5-10% of fine creation and send processes also include an appeal to judge, which is inserted after the fine is sent (frequency = 10-20). This process includes a date appeal to the prefecture, as well as a notification to the offender and potentially a payment process.
* An additional 5% of fine creation and send processes involve a credit collection process, which may be inserted after a payment process or after an appeal to judge (frequency = 16).
* The remaining 20-30% of fine creation and send processes go through other processes, such as a date appeal to the prefecture without a fine sent, a fine sent but with no date appeal to the prefecture, or a fine sent but with no payment or appeal to judge. These processes may also include an insertion of a fine notification, which may also include a payment process or an appeal to judge.

Note that these percentages are just examples and may not accurately reflect the actual performance of the given workflow processes.