Towards the Application of Artificial Intelligence: Cosine Similarity in Recommendation Systems Based on Collaborative Filtering

The development of digital tourism drives the need for a recommendation system in the tourism sector. The need for a recommendation system supported by the development of Artificial Intelligence allows the tourism industry to offer a highly customized experience to customers. One of them is analyzin...

Full description

Saved in:
Bibliographic Details
Published in2024 7th International Conference of Computer and Informatics Engineering (IC2IE) pp. 1 - 5
Main Authors Hartatik, Isnanto, R. Rizal, Warsito, Budi, Firdaus, Nurul, A'La, Fiddin Yusfida
Format Conference Proceeding
LanguageEnglish
Published IEEE 12.09.2024
Subjects
Artificial intelligence
Collaborative filtering
Data models
Filtering
Informatics
Museums
prediction
rating
recommendation
Recommender systems
Three-dimensional displays
Tourism industry
Urban areas
Online AccessGet full text
DOI10.1109/IC2IE63342.2024.10747946

Cover

More Information
Summary:The development of digital tourism drives the need for a recommendation system in the tourism sector. The need for a recommendation system supported by the development of Artificial Intelligence allows the tourism industry to offer a highly customized experience to customers. One of them is analyzing behavioral data and customer interactions so that it can predict the most relevant products, services, or content to recommend. This study aims to develop a tourist destination recommendation model using the cosine similarity method. The Cosine Similarity method is easy to implement and fast to calculate. The dataset used in this study is obtained from 297 users and 57 place tourists. This study has contributed to applying scientific data in the tourism sector by utilizing online platform data to provide 3 cosimilarity-based tourism recommendations where the best model was obtained with an RMSE value at epoch 100 of 0.38. Further research is the use of more complex recommendation methods.
DOI:10.1109/IC2IE63342.2024.10747946