Automatic Mail Rejection Feature

Abstract

A spam defining system defines rules about emails depending on user's reactions to emails. A user can delete an email as spam, or not spam, or without committing to whether the email is spam or not. If the user indicates whether the email is spam or not spam; characteristics of the email are used to update a database. Incoming emails are compared against the database, to determine a likelihood of whether they are spam.

1 . A method, comprising:
 using a computer for obtaining a message;   monitoring a user's actions on said computer with respect to said message;   using the user's actions for automatically forming rules indicating the desirability of said message;   analyzing parts of said message, where said parts include at least at least one hyperlink within said message, and at least information within said message indicative of an internet address within said message; and   said computer using said rules to assess desirability of other messages.   
   
   
       2 . A method as in  claim 1 , wherein said other messages are electronic mail messages. 
   
   
       3 . A method as in  claim 1  wherein said user's actions include a specific way that the user deletes said message using the computer. 
   
   
       4 . A method as in  claim 1  wherein said desirability comprises whether said message is a spam e-mail message. 
   
   
       5 . A method, comprising:
 using a computer for determining a plurality of characteristics of an unwanted electronic message;   using the computer for forming a list with said plurality of characteristics;   using the computer for receiving an incoming electronic message different than said unwanted electronic message, and forming a score of the incoming message by comparing said incoming message with said list and determining commonalities between said incoming message and said list, wherein said comparing comprises determining a domain of the sender, and comparing said domain of the sender with information about spam messages in the list, to obtain a higher probability of spam when information about all senders from a specific domain in said database represent spam, and to represent a lower probability of spam when some senders from said domain in said database represent spam and other senders from said domain in said database do not represent spam;   using the computer for defining said incoming message as likely being spam if said score is within a predetermined range; and   using the computer for taking an action to restrict said message based on said defining.   
   
   
       6 . A method as in  claim 5 , wherein said comparing also comprises determining hyperlinks within said electronic message, and comprises comparing said hyperlinks with hyperlinks within said list. 
   
   
       7 . A computer product, comprising a processor and memory, and executable instructions that are adapted to be executed to implement a filter for an e-mail program, said product comprising:
 an email receiving part that receives information indicative of emails that have been received and determines domain information from a sender of one of said emails;   a storage part that stores information indicative of emails that are known to represent spam, said storage part storing domain information for multiple of said emails that are known to be spam; and   a comparing part, that compares said received emails to said known spam emails, and determines that a received email represents spam responsive to said domain information in said storage part matching to said domain information from said email receiving part.   
   
   
       8 . A product as in  claim 7 , wherein said comparing part determines a received email as being likely to more likely to represent spam when said domain information matches to an uncommon domain, and as being less likely to represent spam with said domain information matches to a less common domain 
   
   
       9 . A product as in  claim 8 , where said comparing part determines a number of hits to the domain to determine whether the domain is a common domain. 
   
   
       10 . A product as in  claim 7 , wherein said comparing part also determining hyperlinks within said electronic message, and comprises comparing said hyperlinks with hyperlinks within said list. 
   
   
       11 . A program as in  claim 7 , further comprising a display output which displays a likelihood of spam coefficient which indicates a numerical percentage likelihood that message specific email represents spam. 
   
   
       12 . A computer product, comprising a processor and memory, and executable instructions that are adapted to be executed to implement a filter for an e-mail program, said product comprising:
 an email receiving part that receives information indicative of emails that have been received and determines information about said emails, one part of said information including information about hyperlinks within said emails that have been received;   a storage part that stores information indicative of emails that are known to represent spam, said storage part storing information about hyperlinks that are within at least one of said emails that are known to be spam; and   a comparing part, that compares said received emails to said known spam emails, and determines that a received email represents spam responsive to said hyperlinks in said storage part matching to said hyperlinks from said email receiving part.   
   
   
       13 . A product as in  claim 12 , wherein said comparing part also determines that a received email represents spam responsive to said domain information in said storage part matching to said domain information from said email receiving part. 
   
   
       14 . A product as in  claim 13 , wherein said comparing part determines a received email as being likely to more likely to represent spam when said domain information matches to an uncommon domain, and as being less likely to represent spam with said domain information matches to a less common domain 
   
   
       15 . A product as in  claim 13 , where said comparing part determines number of hits to the domain to determine whether the domain is a common domain. 
   
   
       16 . A program as in  claim 12 , further comprising a display output which displays a likelihood of spam coefficient which indicates a numerical percentage likelihood that message specific email represents spam.