Research-Based on Telecommunication in Mobile Service Provider's Performance using Enhanced Naive Bayes Classifier

Abstract

In recent years, mobile service providers have rapidly expanded across all countries. Considering unpredictable development trends, mobile service providers are essential to knowledge-based service businesses. Performance may be improved by creating and disseminating new information through innovation activities based on the usage of business intelligence. This research examined the performance of mobile service providers across all countries utilizing an enhanced Naive Bayes classifier based on telecommunication. In comparison to quantitative variables, the naive Bayes performs quite well. In the beginning, data is collected and the normalization technique is used for data preprocessing. Feature extraction is carried out using “Term Frequency and Inverse Document Frequency (TF-IDF)”. “Decision Tree algorithm” is used for data analysis. Then the feature is selected using a two-stage Markov blanket algorithm. Enhanced Naïve Bayes Classifier is the proposed algorithm for telecommunication analysis and at last, the performance of the system is analyzed. This proposed algorithm is used to compare the mobile service provider's performances with existing algorithms. The proposed method measures the following metrics as Throughput, Packet loss, Packet duplication, and User quality of experience. The proposed algorithm is more effective and produces better results.