What’s the difference between user-based and non-user-based filters?
Spam filters come in two varieties: user-based and non-user-based. But I wonder, what exactly does this mean? After all, I’m a computer user, so wouldn’t my spam controls be user based?
Not necessarily. For example, if I worked for a large company and accessed my email on the corporate intranet, I may or may not have user-based spam filtering. The company’s server may do the filtering on my behalf, removing any messages it deems as spam long before they arrive in my mailbox. In this case, this would be a server-based spam filter.
Depending on how the network administrators have set up spam filtering, I may also have a user-based spam filter. If this were the case, the server passes all emails to my inbox. My computer would have a local spam filter which would then handle the messages according to settings that I have set up to meet my individual needs. In this case, this would be a user-based spam filter.
Server-based spam filtering has its pros and cons which is why some companies implement it and others prefer user-based filters. On the plus side, workers don’t have to deal with unwanted junk mail and administrators don’t have to hear complaints about sexually offensive material. Blocking such material is also a proactive move as far as sexual harassment policies go too. However, blocking messages companywide at the server level also has the potential to block legitimate messages. Users do not have control over the level of filtering that takes place, nor do they have the ability to scan the deleted messages in search of legitimate ones.
User-based spam filtering puts the user in control of the filtering. They can adjust the filter to best meet their needs as well as view the Junk Mail folder to see if any legitimate emails have been filtered along with the junk. In addition, they can add or block senders on an individual basis. While this type of filtering is ideal for individual computer users, it’s not necessarily so in a corporate environment. The servers must still handle and route spam traffic which puts a greater load on them and administrators may not want users dealing with spam, especially if there’s a lot of it and if it’s sexually (or otherwise) offensive.
I also want to mention a related type of spam filtering which is similar to user-based filtering: community-based filtering. With community-based spam filtering, the anti-spam solution offers some sort of reporting function. When a message arrives that the user deems as spam, they click a button on the interface to report the message as spam. Likewise, a similar button is used for reporting legitimate messages as “not spam.” If the spam filter marks a legitimate message as spam, you can inform the developers of your opinion.
As a member of a larger community of users, your input is entered into a database. If enough community members report a particular message as spam, it will then become part of the algorithm. For example, spammers send millions of messages at once. A new spam message will get through filters that haven’t yet identified it as spam. However, when enough users say, “This is spam,” this particular message will be identified and blocked.
What about newsletters? What if enough users report this type of email as spam? Subscribers to newsletters may disagree. In community-based environments, this problem is solved when enough users say, “This is not spam.”

