DESIGN AND REALIZATION OF AUTOMATIC DISTRIBUTION REVIEW SYSTEMS BASED ON GAIL-SHAPLE ALGORITHM

Abstract

The article provides an example of implementing the Gale-Shapley algorithm in an automatic article distribution system for peer review. The algorithm, its properties, and application specifics are discussed. The requirements are specified and the database structure is presented, consisting of the tables "articles", "reviewers", "solution_reviewers", "keywords", "reviewer_keywords", "article_keywords". Attention is paid to frameworks, development environments and servers. The functionality of the information system, tested on real data from a scientific publication, is described. Functions related to adding, editing, and deleting articles, working with keywords, and reviewers are detailed. It is noted that the distribution results will vary depending on the initiating party ("reviewer" or "article topic"). It is also established that the distribution will be more advantageous for the initiating party, in this case, ensuring the selection of the most competent reviewer for publication. Additionally, a software product based on this algorithm has been developed. It associates options for selecting reviewers with keywords and prioritizes their assignment. Manual adjustments to the distribution are also possible. A drawback of this system is its focus on mass and large volumes of input information. An advantage is that the reviewer assignment process can be partially automated, especially with a large number of incoming messages. It is demonstrated how the implementation model of the Gale-Shapley algorithm for the problem of distributing articles by reviewers is the basis of a system that optimizes the work of the publication, speeds up and makes individual processes more efficient, in particular the process of reviewing articles taking into account key parameters.