Building a Spam Filter with Naive Bayes (m433)

Hi all,
Upload my project for Naive Bayes spam filter.
Wait criticism, remarks and notices.
Best Regards, Vadim Maklakov.
Building_a_Spam_Filter_with_Naive_Bayes_m433.ipynb (227.8 KB)

Click here to view the jupyter notebook file in a new tab

1 Like

Hi Vadim,

Congratulations on completing another great work and thanks for sharing it with the Community! Well-organized project structure, clean code, excellent code commenting and function documentation, laconic but clear and exhaustive observations throughout the project, cool cover picture.

Some comments from my side:

  • The graph: you can consider despining and de-ticking it, making the title and tick labels bigger, while the figure itself smaller.
  • It’s better to add a short markdown explanation after the code cell [1], at least about the absence of missing values. I know that it’s clear from the output, but it’s always better to mention it explicitly.
  • The code cell [5]. I would add some separators or subheadings for the output, otherwise everything looks attached.
  • A good idea would be to avoid creating the bayes_classify_draft function. You can remove that print statement from the beginning and then just test the function on any string and print out the result.
  • The code cell [13] - it’s better to show the entire messages, not truncated.
  • I would add some markdown explanations after the code cell [14]. Practically, you can cut them from the conclusion. And in the conclusion shortly repeat them. Or even entirely, well, they are not so long :slightly_smiling_face:
  • I’d suggest you to always avoid any new information or examples in the conclusion, especially pieces of code. Only brief mentioning of the main insights of the project.

Hope my suggestions were useful.
Great job as usual! Waiting for your new projects!

1 Like

Elena, thank you very mach for yours notices and suggestions.
I keep its in my mind for future projects.

1 Like