Machine learning approach to automated correction of ETgX documents

The problem is the automatic synthesis of formal correcting rules for LATEX documents. Each document is represented as a syntax tree. Tree node mappings of initial documents to edited documents form the training set, which is used to generate the rules. Rules with a simple structure, which implement removal, insertion or replacing operations of single node and use linear sequence of nodes to select a position are synthesized primarily. The constructed rules are grouped based on the positions of applicability and quality. The rules that use tree-like structure of nodes to select the position are studied. The changes in the quality of the rules during the sequential increase of the training document set are analyzed.