Detection of Multilingual Spam SMS Using NaïveBayes Classifier

The exponential growth of mobile communication has led to an increase in the volume of spam SMS messages, posing a significant challenge to users. This research paper focuses on the detection of multilingual spam SMS using the Naive Bayes classifier. The proposed approach leverages the Naive Bayes algorithm's simplicity and efficiency to classify SMS messages as spam or non-spam, while specifically addressing the multilingual aspect. A comprehensive dataset of labeled SMS messages is collected and preprocessed, including text normalization, and tokenization, and language translation. Feature extraction techniques such as wordfrequency, keyword presence, and message length are employed to represent the SMS messages in a numerical format suitable for the Naive Bayes classifier. The conditional probabilities of features occurring in spam and non-spam messages are calculated using maximum likelihood estimation. To assess the system's performance, extensive testing is conducted, including accuracy assessment, precision computation, recall, and F1 score calculation. The system's ability to handle different languages and accurately detect multilingual spam SMS messages is thoroughly analyzed. The results demonstrate the efficacy of the novel approach in detecting multilingual spam SMS. The Naive Bayes classifier exhibits high accuracy and efficiency in classifying spam messages across various languages, offering users protection from unwanted and potentially harmful content. Furthermore, to facilitate easy utilization and accessibility, a user interface is developed to interact with the spam detection system. The user interface allows users to input messages and the language of the message and receive real-time feedback on the likelihood of a message being spam. The research findings highlight the significance of considering language diversity in spam detection systems and provide insights into the difficulties and opportunities associated with multilingual spam SMS detection. The proposed approach can serve as a groundwork for developing robust and language-aware spam detection systems, ultimately enhancing users' messaging experience and security.