Comparing Two Different Approaches in Big Data and Business Analysis for Churn Prediction with the Focus on How Apache Spark Employed
This work is incremental, as it reviews existing methods for churn prediction in big data contexts, primarily of interest to businesses and data scientists using Spark.
The paper compares two different approaches for churn prediction in business analysis using Apache Spark, focusing on how Spark is utilized to handle big data efficiently, but does not report specific results or numbers.
Due to the significant importance of Big Data analysis, especially in business-related topics such as improving services, finding potential customers, and selecting practical approaches to manage income and expenses, many companies attempt to collaborate with scientists to find how, why, and what they should analysis. In this work, we would like to compare and discuss two different approaches that employed in business analysis topic in Big Data with more consideration on how they utilized Spark. Both studies have investigated Churn Prediction as their case study for their proposed approaches since it is an essential topic in business analysis for companies to recognize a customer intends to leave or stop using their services. Here, we focus on Apache Spark since it has provided several solutions to handle a massive amount of data in recent years efficiently. This feature in Spark makes it one of the most robust candidate tools to upfront with a Big Data problem, particularly time and resource are concerns.