Voyager AI: An Intelligent Content-Based Travel Recommendation System using TF-IDF Vectorization and Cosine Similarity on a Django Full-Stack Architecture

E. Sandhya, Dr. S. Usharani

Content-Based Filtering; TF-IDF Vectorization; Cosine Similarity; Django Web Framework; Personalized Recommendation System.

The proliferation of digital travel platforms has generated immense volumes of destination data, yet existing systems remain inadequate in addressing the nuanced personalization requirements of individual travelers. Generic popularity-based rankings and rigid categorical filters fail to capture the semantic richness of individual travel preferences, thereby constituting a significant unresolved challenge in intelligent information retrieval and personalized recommender system design. This paper presents Voyager AI, a full-stack intelligent travel recommendation system engineered to bridge the gap between unstructured user preference expressions and algorithmically curated destination discovery. The proposed system is implemented using the Python-Django 5.x web framework integrated with a content-based filtering engine built upon Term Frequency-Inverse Document Frequency (TF-IDF) vectorization and cosine similarity computation via the scikit-learn library. The architecture encompasses three principal layers: a relational data layer modelled through Django’s Object-Relational Mapper (ORM) managing Destination, User Preference, and Itinerary entities; a machine learning inference layer that transforms free-form natural language preference descriptions and structured categorical inputs into dense numerical vectors and ranks destinations by semantic alignment; and a responsive presentation layer constructed with Bootstrap 5 and custom CSS implementing a glass morphism design paradigm. A key advantage of the content-based approach is its immunity to the cold-start problem, enabling meaningful personalization from the user’s inaugural session without requiring historical interaction data. The system additionally incorporates a budget-aware post-filtering stage and a complete itinerary management module. Experimental validation using a curated multi-category destination dataset demonstrates accurate semantic retrieval across beach, mountain, historical, adventure, and nature destination categories. The findings establish that open-source NLP techniques, when rigorously integrated within a full-stack web architecture, can deliver recommendation quality commensurate with commercial travel intelligence platforms, whilst remaining computationally accessible for independent deployment.

[1] C. C. Aggarwal, Recommender Systems: The Textbook, Springer International Publishing, 2016.
[2] M. Balabanovic and Y. Shoham, “Fab: Content-Based, Collaborative Recommendation,” Communications of the ACM, Vol. 40, Issue 3, 1997, pp. 66–72.
[3] M. Barreda-Angeles and M. Mateus, “Adoption of Artificial Intelligence in the Travel Industry: An Extended Technology Acceptance Model,” Tourism Management Perspectives, Vol. 37, 2021, p. 100789.
[4] R. Burke, “Hybrid Recommender Systems: Survey and Experiments,” User Modeling and User-Adapted Interaction, Vol. 12, Issue 4, 2002, pp. 331–370.
[5] Django Software Foundation, Django Documentation, Version 5.0, 2024. Available: https://docs.djangoproject.com/en/5.0/
[6] A. Holovaty and J. Kaplan-Moss, The Definitive Guide to Django: Web Development Done Right, 2nd ed., Apress, 2009.
[7] P. Lops, M. de Gemmis, and G. Semeraro, “Content-Based Recommender Systems: State of the Art and Trends,” in Recommender Systems Handbook, Springer, 2011, pp. 73–105.
[8] C. D. Manning, P. Raghavan, and H. Schutze, Introduction to Information Retrieval, Cambridge University Press, 2008.
[9] W. McKinney, “Data Structures for Statistical Computing in Python,” Proceedings of the 9th Python in Science Conference, 2010, pp. 51–56.
[10] F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” Journal of Machine Learning Research, Vol. 12, 2011, pp. 2825–2830.
[11] G. V. Rossum and F. L. Drake, Python 3 Reference Manual, CreateSpace, 2009.
[12] G. Salton and C. Buckley, “Term-Weighting Approaches in Automatic Text Retrieval,” Information Processing and Management, Vol. 24, Issue 5, 1988, pp. 513–523.
[13] J. B. Schafer, D. Frankowski, J. Herlocker, and S. Sen, “Collaborative Filtering Recommender Systems,” in The Adaptive Web, Springer, 2007, pp. 291–324.
[14] P. N. Tan, M. Steinbach, A. Karpatne, and V. Kumar, Introduction to Data Mining, 2nd ed., Pearson Education, 2019.
[15] R. Turrin, M. Quadrana, A. Condorelli, R. Pagano, and P. Cremonesi, “Content-Based Features for Travel Recommendation,” Proceedings of the 8th ACM Conference on Recommender Systems, 2014, pp. 397–398.
[16] H. Werthner and F. Ricci, “E-Commerce and Tourism,” Communications of the ACM, Vol. 47, Issue 12, 2004, pp. 101–105.