A decision support system for assigning reviewer

The support processes of the projects may be negatively affected due to the problems experienced in determining the research areas of R&D projects and determining the appropriate referees for these research fields, together with the recently increasing number of R&D projects and different project fields of activity. In this study, in order to classify R&D projects, the data in the database where the study was carried out were cleaned. Then, "Convolutional Neural Networks (CNN) models have been tried to be created in order to classify features by using the automatic feature learning approach with the word representation method, which is one of the natural language techniques, "Word2Vec". However, it was aimed to measure the reviewers performance in a hierarchical structure. Performance criteria values of the main groups and subgroups were integrated into the VIKOR method with the hesitant fuzzy numbers used in the group multi-criteria decision making approach, and the performances of the R&D project reviewers were tried to be measured according to the TOPSIS method. Finally, an artificial intelligence-based recommendation system has been developed that allows classification of R&D projects with text recognition methods and ranking the assigned of reviewers to these projects using multi-criteria decision support methods.