Log sampling method of RPA workflow, storage medium and equipment

The invention discloses a log sampling method for RPA workflow, which belongs to the field of software engineering, and comprises the following steps of: uniformly analyzing and extracting a log template sequence of each node by acquiring a log of each node on an RPA workflow track; enhancement is carried out through a local flip template sequence, and a positive sample is obtained; carrying out coding and message propagation on the log template sequence based on word embedding and a graph neural network to obtain a graph-level feature expression; the feature expression of the final RPA workflow log is obtained based on a positive sample and graph contrast learning technology, the association information of logs between different nodes is learned through the method, semantic information and structural information in the RPA workflow log are deeply mined, a graph-level portrait is constructed, clustering and inclined sampling are carried out, and the RPA workflow log is obtained. The redundant log quantity is