Scholarlink

Communication is fundamental to written interaction; Humans convey information through messages, ideas, and feelings. Utilizing phones has also emerged using text, sending emails, video, and other forms of interconnections. With the development of technology, there has been a high increase in phone advertisements in the form of texts/emails. The number of devices sometimes becomes flooded by spam messages. These spam messages are misleading and can lead to losing one’s privacy. The most dangerous thing is that one can get. phished and other cyber-attacks on the user through these spam messages. This paper intends to compare and develop techniques for detecting spam and not-spam messages by utilizing Machine Learning algorithms, including the Gaussian Navie Bayes classifier, the Multinomial Navie Bayes classifier, and the Bernoulli Navie Bayes classifier. This is because, with the emergence of new internet users and the possession of private details of many companies, SMS/email spam is prevalent. The results achieved within the study’s scope included an average accuracy of 98% and a nearly perfect precision of 1. When one gets the best model, the model is taken and made into a web app using Streamlit and GitHub, which can be found online.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Spam Detection: A Binary Classification Approach in Machine Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Kunal Bandale,
  • Shital Y. Gaikwad,
  • Ankush D. Sawarkar,
  • Ganesh K. Pakle,
  • Ankit Mahule

摘要

Communication is fundamental to written interaction; Humans convey information through messages, ideas, and feelings. Utilizing phones has also emerged using text, sending emails, video, and other forms of interconnections. With the development of technology, there has been a high increase in phone advertisements in the form of texts/emails. The number of devices sometimes becomes flooded by spam messages. These spam messages are misleading and can lead to losing one’s privacy. The most dangerous thing is that one can get. phished and other cyber-attacks on the user through these spam messages. This paper intends to compare and develop techniques for detecting spam and not-spam messages by utilizing Machine Learning algorithms, including the Gaussian Navie Bayes classifier, the Multinomial Navie Bayes classifier, and the Bernoulli Navie Bayes classifier. This is because, with the emergence of new internet users and the possession of private details of many companies, SMS/email spam is prevalent. The results achieved within the study’s scope included an average accuracy of 98% and a nearly perfect precision of 1. When one gets the best model, the model is taken and made into a web app using Streamlit and GitHub, which can be found online.