Further Work
- In U4, Brian and I agreed to move a page very similar to this to the U4 TG, and I suggest we move this one to the U3 TG too. --MF
- Could this be TIF content appended to the previous page? --MF
-
What else could you do to improve the spam filtering? Discuss this with your team. You might consider the following options:
- Detect spam by finding phone numbers or web links
- Detect spam by finding words that are highly frequent in spam messages and not in ham messages
- Detect spam by checking with a dictionary for clean spelling and no abbreviations
- Combine several methods
- How hard would it be to implement your solutions?
- Compare your solutions with the one you wrote earlier (using average number of words)? Which would do better? Why?