Online relational databases are quite complex to analyze, particularly dynamically. The performance of the database for add, delete, and update is important. Various strategies for quick updating of databases are being studied by different researchers.

This paper is an important contribution toward solving such problems. The authors consider online dynamic transactions. Normal forms of transactions (similar to the normal forms of relations) are proposed. A number of axioms, lemmas, and theorems are presented to show the minimality of transactions while simplifying the transactions for update. The concept of a procedure forest as a tool in proving various results is an important contribution to this field. It may also help in the design of parallel algorithms for transaction updates, which are important, considering the advances in parallel architecture compilers and their expected use.

I strongly recommend the paper to researchers and graduate students in the theoretical aspects of relational databases and their use in transaction processing. This work will also interest researchers working in parallel algorithms. Computer scientists involved in the development and improvement of database software may find the theoretical ideas of the paper important in improving the performance of a relational database during updates, particularly in SQL environments. Considering the length of the paper and the importance of the topic, some more relevant references would have helped the reader.