@ARTICLE{Alesheikh, author = {neisani samani, Zeinab and Alesheikh, Ali Asghar and Sadeghi-Niaraki, Abolghasem and Nazari Ashani, Mahdi and }, title = {Personalization of a tourism recommender system based on users similarity and the use of deep belief network}, volume = {10}, number = {4}, abstract ={Spatial recommendation systems allow users to provide useful information by reducing duplicate and irrelevant information on the web widely. Recommendation systems are widely used in various fields, including tourism. Tourism recommendation systems can be used as tools by the tourist. A tourist can visit the tourist attractions of the region in a short time and with the least facilities, cost, and knowledge. Recommendation systems generally offer the necessary suggestions to different users based on participatory refinement and the similarity between the groups of the users. However, many services do not match the personal characteristics of the individual, and this reduces the effectiveness of such systems. The purpose of this study is to develop a recommendation algorithm based on the similarities between the users and personalization concepts. The innovation of this research is the use of a deep belief neural network to personalize the suggestions that can be offered to the tourists. The research scenario is as follows: first different tourists register in the system; then they express their personal information and general preferences and specific personalization factors for visiting the tourist centres. In the proposed approach, there is no need to separate the users; rather, due to the learning power of deep neural networks, it is possible to differentiate and personalize the user suggestions. In this regard, the related data to 400 tourists were received based on 14 input and distinguishing elements. Furthermore, based on the trained network, the predictability of personalized tourist places for 30 new users was examined. The results were evaluated based on these three indicators: Precision, Recall, and F-Score, as well as the user satisfaction. The results showed the high accuracy as well as the satisfaction of more than 79% of the users. }, URL = {http://jgit.kntu.ac.ir/article-1-825-en.html}, eprint = {http://jgit.kntu.ac.ir/article-1-825-en.pdf}, journal = {Journal of Geospatial Information Technology}, doi = {10.61186/jgit.10.4.1}, year = {2023} }