Personalised Location Based Recommendation System Using Apache Spark
Vara Swetha, S. Zeenath Fathima, Tungala Tharuni · 2023
Recommender systems play a very crucial role in our lives by suggesting the items users might be interested in without much provision from them. In recent times, the spike in internet usage and advancement in the Global Positioning System (GPS) has opened new avenues to explore the already thriving area of Location-based Services and provide recommendations based on them. The location-based recommendation has been lauded for its invaluable solution to information overload on the internet. This type of recommendation system aims to dispense service recommendations and product recommendations based on the profile as well as preferences of the users. Recommendation systems can be built in many ways and here, the aim is to do it using the PySpark framework that employs data from several sources including GPS, user reviews, and location-based searches to prove the recommendations. PySpark is used for managing the large datasets. In this paper first, we executed the ML algorithm which is the k-means algorithm and after that, we used the spark for the execution of the dataset where we used two algorithms ALS and SVD ++, and achieved an RMSE of 1.5 and 1.3 respectively