Proof of Concept: Anti-Spam FilterFiled Under: Programming, Projects
I have been working on a anti-spam filter for comments to use on my website arthurwinters.net. This filter takes as string of text, analyzes it and returns either a 0 (not spam) or a 1 (spam). The spam filter is based on a scoring system which judges the following:
- Word Count and Character Count.
- How many URLs are in the post.
- How many bad domain keywords, and path keywords in the post.
- How many bad keywords are in the post.
- How many constant characters are in a row.
Depending on the situation it could be given a positive score or a negative score. This combined with an overall score will give the post a percentage, based on that percentage and a threshold determined, it is then marked as spam or not spam.
Feel free to try it out at http://www.arthurwinters.net/beta
- Permalink
- Jinan
- 1 Nov 2008 10:12 PM
- Comments (1)
November 30th, 2008 at 4:01 pm
Or you could just use Akismet!