Spam Detection

Comparison and analysis of several algorithms on the problem of detecting spam in messages. The analysis of the data set is carried out.

• SMS Spam Collection Dataset
• The files contain one message per line. Each line is composed by two columns: v1 contains the label (ham or spam) and v2 contains the raw text.

Realised methods:

• Naive Bayes spam filtering
• K-Nearest Neighbors algorithm
• Decision Tree learning
• Support Vector Machine (SVM)
• Random Forest

Methods

1. Naive Bayes spam filtering

• Naive Bayes spam filtering
• Naive Bayes classifier
• Text Classification and Naïve Bayes
• Naive Bayes - Stanford NLP - Professor Dan Jurafsky

2. K-Nearest Neighbors algorithm

• Description
• Nearest Neighbors

3. Decision tree learning

• Description #1
• Description #2

4. Support Vector Machine (SVM)

• Description

5. Random Forest

• Description

Libraries

• WordCloud

  • https://www.datacamp.com/community/tutorials/wordcloud-python

• Seaborn

  • https://seaborn.pydata.org

Else

• Confusion matrix
• Supervised learning