This repository contains a simple spam classification project using the Naive Bayes algorithm. The project aims to classify text messages as spam or ham (non-spam). The dataset used is the SMS Spam Collection Dataset from the UCI Machine Learning Repository.
The dataset is explored by performing text cleaning, removing duplicates, and analyzing the data.
- Pie chart: Visualizes the distribution of spam and ham messages.
- Bar chart: Illustrates the count of spam and ham messages.
- Word cloud: Displays a word cloud representation of spam and ham messages.
Tokenization is performed to break the text into sentences, words, and characters.
- Sentence tokenization: Splits the text into sentences.
- Word tokenization: Divides the text into individual words.
- Character tokenization: Breaks the text into individual characters.
The ham and spam categories are described, providing insights into their characteristics and composition.
Several pretrained classifiers are employed to classify the spam messages.
| Model Name | Accuracy | Time Taken |
|---|---|---|
| AdaBoostClassifier | 96.611810 | 0.696584 |
| BaggingClassifier | 95.934172 | 1.062076 |
| BernoulliNB | 97.386254 | 0.012100 |
| CalibratedClassifierCV | 98.644724 | 0.177198 |
| ComplementNB | 96.515005 | 0.010757 |
| DecisionTreeClassifier | 95.643756 | 0.131743 |
| DummyClassifier | 88.770571 | 0.001799 |
| ExtraTreeClassifier | 96.127783 | 0.017295 |
| ExtraTreesClassifier | 98.160697 | 1.323269 |
| GradientBoostingClassifier | 96.708616 | 1.091855 |
| KNeighborsClassifier | 92.545983 | 0.173880 |
| LogisticRegression | 97.676670 | 0.058000 |
| LogisticRegressionCV | 98.451113 | 2.000187 |
| MLPClassifier | 98.063892 | 32.597064 |
| MultinomialNB | 98.160697 | 0.040769 |
| RandomForestClassifier | 97.386254 | 1.374330 |
| SGDClassifier | 97.773475 | 0.009635 |
| SVC | 97.483059 | 1.680027 |
Accuracy score: 0.9796708615682478
Accuracy score: 0.9864472410454985
Accuracy score: 0.9002904162633107
Accuracy score: 0.8993223620522749
Please note that this project was developed by a college student during late-night hours, and the code might not be extensively documented. However, feel free to explore and modify the code as per your requirements. If you encounter any issues or have suggestions for improvement, please submit an issue.
Contributions are welcome! If you'd like to contribute to this project, please feel free to submit a pull request with your proposed changes.
This project is licensed under the MIT License. See the LICENSE file for more information.
Enjoy exploring the spam classification project and have a great time!