Feature Analysis and Effectiveness

Allow/Block List:

 

Effectiveness: High

 

Advantages: Allows certain messages that have a word, phrase, email address, domain, or ip address that appear in the header or body of the message to bypass SpamBully when you add them to your Allow list and make it to your Inbox.  Messages containing words, phrases, email addresses, domains, or ip addresses can be blocked automatically as well when you add them to your Block list and will be sent to the Spam folder.

 

How it works:  After you have added a word, phrase, domain or email address to your Allow or Block list, any message containing that word/phrase will be allowed to pass through SpamBully if added to your Allow list or blocked if added to your Block list.  This can decrease false positives and increase the accuracy of classifying spam. You can also add the wild card character * to any entry to handle basic pattern matching. For example adding *@domain.com as an email address to your allow list would allow any email from domain.com to automatically bypass the filter.  Your Contacts in Outlook or Outlook Express, as well as any users you have replied to in the past are automatically considered as friends by default and allowed to pass through the filter and sent to your Inbox.  

 

Limitations: Word phrases to allow or block messages must be carefully chosen so that only the proper messages are allowed or blocked by the word or phrase. If you pick a common word which also appears in spam messages to your Allow list you may let through unnecessary spam that normally would have been blocked.

 

Recommendations:  We recommend enabling this filter for almost all users.  Be careful with the word phrases you put in your allow list so you don't accidentally allow spam messages through.

 

 

Bayesian Filter:

 

Effectiveness: High

 

Advantages:  The heart of Spam Bully; trained properly the Bayesian filter can be as effective as you are in filtering spam messages.

 

How it works:  Bayesian filters work by looking at the words that are used in spam emails and good emails and then classifying the email based on the words that these emails contain.  There are certain words that tend to comprise spam emails and also certain words that tend to comprise your good emails.  By analyzing these patterns SpamBully can effectively route spam messages to the Spam folder and good emails to your Inbox.  It can adapt itself to your email as you correct any mistakes it may make.  SpamBully 3 also further analyzes email messages by looking up the originating country and language of the message.

 

Limitations: Bayesian filters that are improperly trained can be inaccurate.  It can also take up to a few weeks for Bayesian filters to become fully trained.  It is recommended that you closely monitor SpamBully during the first few weeks of operation to make sure it is categorizing your email properly. You can correct this by using the Spam and Not Spam buttons in the SpamBully toolbar so SpamBully will better learn from your email habits.

 

Recommendations:  We recommend enabling this filter for almost all users.  Properly trained Bayesian filters can become over 99% accurate in some cases.  One other recommendation we have is to NOT delete all of your old spam messages.  If you ever need to retrain your Bayesian filter, your old spam messages are a goldmine of knowledge to build your Bayesian filter from and can help make your filter even more accurate.

 

Bounce

 

Effectiveness: Very Low

 

Advantages: Bounce is one of the only methods that can actually lower the physical amount of spam you receive. (However there is an equal chance it can increase the level of spam you receive.)

 

How does it work:  When a message is bounced spammers who receive a bounce may remove your email address from their spam lists.  It is most effective against the semi-legitimate spammers who use real email addresses or domain names which will be more likely to receive a bounce message.  This feature is no longer considered an effective strategy against stopping spam in many cases.  The main reason is because most spammers use disposable email addresses and hacked machines.  So they will never receive the bounce.The bounce you send a spammer may bounce right back to you.

 

Limitations: Bouncing all your messages is labor intensive, it requires you to review and bounce the spam messages you receive.  It is ineffective against many spammers as they do not use legitimate return email addresses, so they will never receive the bounces and they may bounce right back to you.  It also uses extra bandwidth to send and retrieve the bounce messages.  Another problem is many of the messages you bounce will be returned to you if the email address the spammer uses is invalid or no longer active.  This could give the appearance that you are receiving more spam when actually you are just getting back the bounce message you sent from SpamBully.

 

Recommendations:  Bouncing is most effective for users with very light amounts of spam messages.  If you receive 5-10 spam messages or less a day, this may be an effective part of your spam solution provided you review each message to determine whether the user is likely to receive the bounce.  For higher levels of spam, generally the effort required to bounce every message is not worth the few lists you may be removed from.  It is more effective to concentrate your efforts on keeping your Bayesian spam filters properly trained.  If you feel you want to get back at spammers, use the report as spam option in message details.  This sends a letter of complaint to the FTC, the spammers ISP and advertisers to hopefully get them banned or investigated.  This may hurt a spammer far more than bouncing a message.

 

 

Challenge Email:

 

Effectiveness: Medium

 

Advantages: Can almost completely eliminate spam.

 

How it works: Is an email that is sent to an unknown user that has passed through the Bayesian filter but is not yet in your Allow list or Contacts.  This email email asks them to confirm they are a valid user by typing in a password from the Challenge Email.  

 

Limitations:  Can increase false positives significantly.  It can also cause problems with net-etiquette.  It is generally not polite and can also be considered unprofessional if using this method with business email accounts.  DO NOT USE this feature if you belong to newsletters or group discussions where you participate with other users.  This can cause a Challenge Email to be sent to all users participating in a newsletter and this type of internet behavior is unacceptable.

 

Recommendations:  We only recommend enabling this if the current filter methods are not effective and you do not belong to any group style participation newsletters.  Please remember that not all users receiving these notices will respond even if they are good users.  Also, all email receipts for goods and services you may have purchased may be blocked when using this feature.

 

Bypass IE Favorites domains:

 

Effectiveness: Medium

 

Advantages:  Reduces the amount of false positives from SpamBully.

 

How it works: Will automatically allow emails from domains that appear in your Internet Explorer Favorites list.  However, if at least one email address from one of these domains has been added to your Block list, the domain will not be allowed to bypass the filter.  This helps to prevent common domains which are used by spammers, to not bypass the filter once an email address has been added to the Block list.

 

Limitations:  Can increase the amount of spam messages slightly particularly when first starting to use this filter option.

 

Recommendations:  We recommend enabling this option if you are seeing too many false positives.  It will cause an initial increase in spam initially until users from common domains are added to your Block list.  These common domains will then no longer be able to use this filter bypass feature.

 

 Bypass IE History Domains:

 

Effectiveness: Medium

 

Advantages:  Reduces the amount of false positives from SpamBully.

 

How it works: Will automatically allow emails from domains that appear in your Internet Explorer History list into your Inbox.  However, if at least one email address from one of these domains has been added to your Block list, the domain will not be allowed to bypass the filter.  This helps to prevent common domains which are used by spammers, to not bypass the filter once an email address has been added to the Block list.

 

Limitations:  Can increase the amount of spam messages slightly particularly when first starting to use this filter option.

 

Recommendations:  Enabling this option if you are seeing too many false positives.  It will cause an initial increase in spam initially until users from common domains are added to your Block list.  After that, these common domains will no longer be able to use this filter bypass feature. 

 

Punish

 Effectiveness: Low

 

Advantages:  Potential of increasing a spammers cost of sending spam.

 

Limitations: It is possible that the spammer will think the hits are legitimate and if the spammer has user level tracking in the links it could make the spammer think he has a person interested in their spam message which could ultimately lead to more spam messages. 

 

How does it work:  Punish loads the web pages in a spam email message a certain number of times.  Since nothing is being bought from these visits it could minimally increase the bandwidth costs to the spammer without any increase in revenue.

 

Recommendations: We don't recommend using punish at this time for most users.  It can be minimally effective at increasing costs to spammers but generally this is not worth the time it takes to punish a message.

  

RBL Servers:

 

Effectiveness: Medium

 

Advantages:  Can block messages coming from known servers that transmit spam. 

 

How it works:  There are lists available on the internet that list open relay servers.  These types of servers allow any user to send messages through them.  It is highly likely that spammers will send messages through these servers and many of emails coming from this server are spam.

 

Limitations: There have been several cases of legitimate email servers being added to these lists and some of these servers also pass legitimate email messages.  So there is a chance that a message you want to receive could be blocked by these RBL lists. Using this filter can slow down SpamBully processing as it must check each message IP address at several remote servers.

 

Recommendations:  If you are having a problem with too many spam messages passing through SpamBully and you have a high speed internet connection, you can enable this filter to help filter more spam.  If you have a slow connection, you can test it for speed and effectiveness in filtering your spam.

 

Report as Spam:

 

Effectiveness: Medium

 

Advantages: Report as spam is one of the only methods to potentially lower the physical amount of spam you receive.

 

How does it work:  Report as Spam looks up the information from the spammers message and finds out who owns the domain and hosts the websites the spammer is promoting. It can also look up the advertisers as well.  Once SpamBully has this information it reports the message to the proper servers.  It also sends a message to the FTC notifying them of the spam message.  The FTC can then decide whether to take further steps against the spammer.

 

Limitations: Reporting all your messages is labor intensive, it requires you to review and report the spam messages you receive one at a time.  This is because it take several seconds to analyze the email and retrieve the proper records from the internet.  We also require messages to be reported one at a time so they can be reviewed by an actual human.  It is serious business reporting a spam message and can have serious consequences for the person doing the spamming.  Therefore, we want each user to review each message before reporting it.

 

Recommendations:  Reporting is most effective for users with very light amounts of spam messages.  If you receive 5-10 spam messages or less a day, this may be an effective part of your spam solution.  For higher levels of spam, generally the effort required to report every message is not worth the time it takes.  It is more effective to concentrate your efforts on keeping your Bayesian spam filters properly trained.  The goal of a spam filter is to spend less time dealing with spam in the first place so reporting each spammer usually means you are spending more and not less time dealing with spam. 

 

Web Bayesian filter:

 

Effectiveness: Medium

 

Advantages:  Can effectively filter messages that may only contain an html link.  This has the potential to catch more spam messages.

 

How it works: Performs a Bayesian analysis of the links in your email messages. Spam Bully intelligently reviews the content on the web pages these email links point to so it can determine if the email is spam or not spam.  This is helpful in analyzing messages that may only contain a link to a website. 

 

Limitations: This process can be slow as each message must be scanned for links and then the web page or pages must be retrieved and processed.

 

Recommendations:  We recommend enabling this feature only if SpamBully is not effectively filtering your messages.  You should also have a high speed connection.  If you have a slow connection, you can test it for speed and effectiveness at reducing your spam.

 


This article may be found at:
http://www.spambully.com/sb3help/index.php?page=index_v2&id=41&c=2



Article Comments  

There are no user added comments for this article.