Experimenting with facilitating collaborative travel recommendations

Abstract

© 2019 IEEE. The aim of undertaken work was to experiment with, and compare, various approaches to facilitate a collaborative recommender system, for travelers who would like to visit 'tourist places'. For this purpose, different algorithms, including Kohonen-type neural networks (plain and elastic), as well as semantic technologies, have been applied to a dataset collected form the Web. Experiments have been performed for groups of users (measuring quality of recommendations), as well as for selected individuals. Completed comparison points out to main strength and weakness of each approach.