Intelligent Hotel Recommendation Engine Using Lstm And Hippopotamus-Inspired Hyperparameter Tuning

Abstract

Recommendation system has gained a lot of popularity in almost all the fields. Hospitality industry has transformed with the gain in popularity of recommendation system. Hotel industry has boomed a lot, as users are getting ample recommendation options depending on their preferences and choices. In this study we have used Long short Term Memory model (LSTM)for hotel recommendations and to increase the accuracy of recommended hotels we have used Hippopotamus Optimization Algorithm (HOA) based on user’s personal preferences. LSTM analyse user’s reviews to understand their preferences for making personalized recommendations. HOA is used to fine tune the hyperparameters of LSTM to give better personalized recommendation which matches user’s choice. This hybrid model gives training accuracy of 97.69% and 0.2830 loss. Validation accuracy of 93.21% and loss of 0.3016. HOA+LSTM model beats classic optimisation techniques, offering a more resilient and reliable recommendation system. This research presents a sophisticated optimisation method that enhances decision-making for users in intelligent tourism.