Distributed accounting affair processing method and system

The invention relates to a distributed accounting affair processing method, which comprises: performing machine learning based on historical information of transaction data stored in a plurality of databases to form an allocation strategy knowledge base, wherein the transaction data comprises a plurality of transaction records; and executing at least one of the following items based on the allocation strategy knowledge selecting one of a plurality of databases to record a new transaction record, newly setting a database to record a new transaction record, and changing a destination database corresponding to stored transaction records. The method is beneficial to enabling the number of accounts managed by each database and the data volume to be relatively balanced. 本发明涉及一种分布式账务处理方法，包括：基于交易数据被存储于多个数据库的历史信息进行机器学习，以形成分配策略知识库；其中，交易数据包括多条交易记录；基于分配策略知识库来执行如下项中的至少一项：选择多个数据库其中之一以记录新的交易记录；新设数据库以记录新的交易记录；更改对应于已存储交易记录的目的地数据库。该方法有利于使各个数据库管理的账户数量以及数据量相对均衡。