Systems and methods for spam filtering
Systems and methods to reduce false positives in spam filtering. In one aspect, a method includes automatically generating profiles for persons identified in messages, such as incoming and outgoing emails. Messages flagged as spam can be checked against the profile to identify false positives.
1. A computer implemented method, comprising:
scanning, by a computer system, a plurality of sent or received messages of a user to identify addresses of a plurality of persons;
extracting, by the computer system, a portion of information from headers of each message, said portion information extracted from the headers corresponding to a first portion of profile data for the persons from the messages;
further extracting, by the computing system, a second portion of profile data for the persons from the messages from online sources separate from the messages;
storing the addresses of the plurality of persons in a plurality of profiles, each profile being for a respective one of the plurality of persons;
adding the extracted first and second portions of profile data to the profiles, wherein the extracted first and second portions of profile data includes information related to names of the persons and phone numbers for the persons that was respectively retrieved from the messages and separate online sources;
querying, by the computer system using the extracted first and second portions of profile data as search criteria, a server in communication with the computer system over a network to obtain additional information regarding the persons;
adding the additional information to the profiles;
updating, by the computer system, the plurality of profiles as new messages of the user are received;
scanning, by the computer system, a set of messages that have been previously classified by a spam filter as spam;
determining, by the computer system, one or more false positive spam messages from the set of messages classified as spam by comparing addresses and names used in the set of messages with addresses of the plurality of persons and names in the plurality of profiles; and
presenting, by the computer system, a list to identify the one or more false positive spam messages to the user.
2. The method of claim 1 , wherein presenting the list includes displaying the set of messages that are classified as spam and highlighting the one or more false positive spam messages within the list.
3. The method of claim 1 , further comprising moving the one or more false positive spam messages out of the set of messages that are classified as spam.
4. The method of claim 3 , wherein the false positive spam messages are moved in response to a single command from a user of the computer.
5. The method of claim 4 , wherein presenting the list comprises highlighting the one or more false positive spam messages in a list of incoming messages.
6. The method of claim 3 , wherein the false positive spam messages are moved in response to the computer receiving the one or more messages and determining that the one or more messages are false positive in spam filtering.
7. The method of claim 1 , wherein each of the sent or received messages is one of: an email, an instant message, and a text message.
8. The method of claim 1 , wherein the computer system is remote to a user terminal on which the list is displayed.
9. The method of claim 1 , wherein the computer system is a user terminal configured to receive incoming messages for the user, send outgoing messages from the user, and display the list.
10. The method of claim 1 , wherein the server includes one or more of: a web search engine, an online directory, a social network server, a business website, a personal website, a media sharing website, a map and direction website, an online retailer, a travel website, and a location website.
11. The method of claim 1 , wherein determining the one or more false positive spam messages includes determining a relevancy score for each of the profiles, and using the relevancy scores of the profiles to filter the spam messages.
12. The method of claim 11 , wherein determining the relevancy score includes determining a relevancy score for each respective address in the plurality of profiles based on a number of previously sent or received messages in which each respective address is used and types of fields in which the address is used in previously sent or received messages.
13. The method of claim 12 , wherein determining the relevancy score includes determining a relevancy score for each respective person associated with the plurality of profiles based on the relevancy score for each respective address for the respective person.
14. The method of claim 1 , wherein determining the one or more false positive spam messages includes determining that one or more addresses used in the one or more false positive spam messages are included in the stored addresses.
15. The method of claim 1 , wherein the scanning of the plurality of messages comprises scanning both sent and received messages to identify the addresses.
16. The method of claim 15 , wherein the received messages and the sent messages are in a first account of the user; the set of messages classified as spam are in a second account of the user; the first account is associated with a first address of the user; and the second account is associated with a second address of the user.
17. The method of claim 1 , wherein determining of the one or more false positive spam messages comprises communicating via a lightweight directory access protocol to identify the one or more false positive spam messages that have addresses listed on a server.
18. A system comprising:
a processor; and
memory coupled to the processor and storing instructions that, when executed by the processor, cause the system to:
scan a plurality of sent or received messages of a user to identify addresses of a plurality of persons;
extract a portion of information from headers of each message, said portion information extracted from the headers corresponding to a first portion of profile data for the persons from the messages;
extract a second portion of profile data for the persons from the messages from online sources separate from the messages;
store the addresses of the plurality of persons in a plurality of profiles, each profile being for a respective one of the plurality of persons;
add the extracted first and second portions of profile data to the profiles, wherein the extracted first and second portions of profile data includes information related to names of the persons and phone numbers for the persons that was respectively retrieved from the messages and separate online sources;
query, using the extracted first and second portions of profile data as search criteria, a server in communication with the computer system over a network to obtain additional information regarding the persons;
add the additional information to the profiles;
update the plurality of profiles as new messages of the user are received;
scan a set of messages that have been previously classified by a spam filter as spam;
determine one or more false positive spam messages from the set of messages classified as spam by comparing addresses and names used in the set of messages with addresses of the plurality of persons and names in the plurality of profiles; and
present, to the user, a list to identify the one or more false positive spam messages.
19. A non-transitory machine readable storage media storing a set of instructions, that, when executed by a computer system, cause the computer system to:
scan a plurality of sent or received messages of a user to identify addresses of a plurality of persons;
extract a portion of information from headers of each message, said portion information extracted from the headers corresponding to a first portion of profile data for the persons from the messages;
extract a second portion of profile data for the persons from the messages from online sources separate from the messages;
store the addresses of the plurality of persons in a plurality of profiles, each profile being for a respective one of the plurality of persons;
add the extracted first and second portions of profile data to the profiles, wherein the extracted first and second portions of profile data includes information related to names of the persons and phone numbers for the persons that was respectively retrieved from the messages and separate online sources;
query, using the extracted first and second portions of profile data as search criteria, a server in communication with the computer system over a network to obtain additional information regarding the persons;
add the additional information to the profiles;
update, by the computer system, the plurality of profiles as new messages of the user are received;
scan, by the computer system, a set of messages that have been previously classified by a spam filter as spam;
determine one or more false positive spam messages from the set of messages classified as spam by comparing addresses and names used in the set of messages with addresses of the plurality of persons and names in the plurality of profiles; and
present, to the user, a list to identify the one or more false positive spam messages.