Description

For starters in the classification of spam sms by importing libraries to be used such as pandas, numpy, matplotlib pyplot, pickle, sastrawi and nltk. Which can be seen as follows :

To make it easier to analyze the contents of the data used, data visualization is carried out. Based on the data visualization carried out, there are 569 normal sms data, 335 fraud or fraud sms data, and 239 promo sms data.

After importing the library to be used and data visualization, the next step is the pre-procesing stage of the text data. Data preprocessing is a technique applied to databases to remove noise, missing values, errors, unimportant data and inconsistent data. The purpose of preprocessing data is to transform raw data for analysis in order to produce quality and accurate data.

The fourth stage is to model data using machine learning. Data Modeling is a process used to determine and analyze the data needs needed to support classification. Ensure that all data objects required by the database are accurately represented. Data omission will lead to incorrect report generation and result in incorrect results.

The final stage is the evaluation of the model. An evaluation model is used to find out how far the success of the program is so that steps are obtained to make improvements or developments.