Recommendation System for Food Dishes in Specific Restaurants Based on Sentiment Analysis

Pal, Sunanda (2023) Recommendation System for Food Dishes in Specific Restaurants Based on Sentiment Analysis. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (814kB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

In the food and restaurant industry, recommendation systems are immensely used to enhance the customer experience. Customers often leave comments to share their dining experiences, either on dedicated food review platforms or on social media. Researchers collect massive amounts of such relevant data to build advanced recommendation systems and improve upon the existing ones. Current research, however, primarily focuses on recommending restaurants based on client preferences or dishes based on higher restaurant reviews. Thus, in this study, a new kind of recommendation system is proposed that suggests certain dishes in a specific restaurant to customers based on their culinary preferences and similarity with other users. To accomplish this, first, NLP methods are used to extract dishes from reviews. Then, sentiment analysis is applied in order to detect which users like which dish and restaurant. Finally, collaborative filtering is used to recommend food and restaurants to the users. Due to the large volume of datasets and available resource restrictions, the model implementation is carried out in two ways. In the first approach, KNN-based algorithms (KNNBasic and KNNwithMeans) are trained on the sampled data, where hyperparameter-tuned KNNBasic with cross-validation has shown better accuracy. In the second approach, matrix factorization techniques are implemented on the entire data, where optimized SVD has performed better than NMF in terms of RMSE and MAE. The research also identifies the limitations of the proposed methodology and provides directions towards the scope of improvements.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Horta, Vitor UNSPECIFIED
Uncontrolled Keywords:	Natural language processing; text-mining; food recommendation; aspect-based sentiment analysis; restaurant-feedbacks; collaborative filtering
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Hospitality Industry > Food service
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	28 Dec 2024 11:20
Last Modified:	28 Dec 2024 11:20
URI:	https://norma.ncirl.ie/id/eprint/7242

Actions (login required)

View Item