aalto1 untyped-item.component.html
You might also like: A comparative analysis of three recommendation algorithms
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Insinööritieteiden korkeakoulu |
Bachelor's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
Department
Major/Subject
Mcode
ENG3082
Degree programme
Language
en
Pages
26
Series
Abstract
The Slush matchmaking platform aims to facilitate meaningful connections between attendees at a large-scale technology event, facing challenges such as a limited operational timeframe and unique user-company matching requirements. This thesis addresses the need for an effective recommendation system within this context by conducting a comparative analysis of three prominent recommendation algorithms: collaborative filtering via Matrix Factorisation with Alternating Least Squares (MF-ALS), content-based TF-IDF, and a hybrid Two-Tower neural network model. Utilising a curated dataset comprising of user profiles, company details, and interaction data from Slush, the study employed a dual-phased evaluation methodology. This included offline testing using ranking metrics and a simulated user testing where participants rated recommendation relevance. The results consistently demonstrate that the Two-Tower hybrid model significantly outperforms both TF-IDF and MF-ALS in accurately retrieving and ranking relevant companies, as validated by both offline metrics and user ratings. The findings conclude that while hybrid approaches offer the most promising accuracy for event matchmaking, practical implementation requires careful consideration of the accuracy-latency balance.