F'ailteFinder: Recommender System implementing NLP and KNN for the Irish tourist domain

Estrada Ventura, Vladimir (2025) F'ailteFinder: Recommender System implementing NLP and KNN for the Irish tourist domain. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (757kB) | Preview

Abstract

F'ailteFinder is a domain-specific recommender system for the Irish tourism sector that delivers personalised suggestions for attractions, events, and accommodation. The system follows CRISP-DM end-to-end: annually refreshed datasets from the F´ailte Ireland API are cleaned and normalised, loaded into PostgreSQL, and exposed to a modular pipeline that includes NLP preprocessing, Word2Vec semantic embeddings, and k-Nearest Neighbours ranking under cosine similarity. Traveller preferences and geospatial proximity are incorporated to address choice overload, limited personalisation, and location-agnostic results common in generic systems. A Streamlit front end operationalises the stack, providing an interactive interface for interests, dates, and accommodation context. The approach demonstrates the potential of content-based filtering enhanced with NLP techniques to transform destination discovery, particularly in niche markets. This research contributes to the academic discourse on recommender systems by presenting a tailored solution for tourism, while offering practical implications for boosting visitor engagement in Ireland. Case studies indicate excellent early precision and low latency, supporting the practical viability of content-based, location-aware recommendations in this domain; remaining gaps are chiefly in list ordering and spatial compactness. Overall, the work demonstrates a maintainable, reproducible architecture that adapts results to individual interests while keeping data current. Future extensions will focus on lightweight distance-aware re-ranking and diversification to refine ordering and variety, with selective exploration of hybrid signals and containerised deployment to strengthen robustness and scalability.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Yaqoob, Abid UNSPECIFIED
Subjects:	P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing D History General and Old World > DA Great Britain > Ireland H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Tourism Industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	25 Aug 2026 13:49
Last Modified:	25 Aug 2026 13:49
URI:	https://norma.ncirl.ie/id/eprint/9629

Actions (login required)

View Item