SYSTEM AND METHOD FOR AUTOMATICALLY GENERATING EMAIL SUBJECT LINES

• US 20150058428A1
• Filed: 09/16/2013
• Published: 02/26/2015
• Est. Priority Date: 08/23/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Caskey, Sasha P., Farrell, Robert G., Kanevsky, Dimitri, Sainath, Tara N.
• Granted Patent: US 9,356,889 B2
• Time in Patent Office: Days
• US Class Current: 709/206
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
• CPC Class Codes: H04L 51/42 Mailbox-related aspects, e.g. synchronisation of mailboxes

Abstract

Systems and methods for automatically generating e-mail subject lines are provided. A method for automatic generation of subject lines for electronic mail (email), comprises extracting topics from an email message, computing a sender relevance score for each topic, computing a receiver relevance score for each topic, ranking the topics based on the sender and receiver relevance scores, extracting syntactic units from the email message, ranking the extracted syntactic units from the email message based on the topic ranking, and assigning one or more subject lines to the email message based on the ranking of the syntactic units.

Claims

1. A method for automatic generation of subject lines for electronic mail (email), the method comprising:

extracting topics from an email message;

computing a sender relevance score for each topic;

ranking the topics based on the sender relevance scores;

ranking a plurality of syntactic units from the email message based on the topic ranking; and

assigning one or more subject lines to the email message based on the ranking of the syntactic units.

2. The method according to claim 1, wherein computing the sender relevance score comprises comparing a term frequency score for a topic to a corpus frequency score for the topic, wherein the corpus frequency score is based on a corpus of emails that have been at least one of sent or received by a sender of the email message over a period of time.

3. The method according to claim 2, wherein computing the sender relevance score further comprises identifying topics that match one or more characteristics of the sender of the email message.

4. The method according to claim 3, wherein the one or more characteristics are derived from behavioral data of the sender taken in connection with the corpus of emails.

5. The method according to claim 1, further comprising:

computing a receiver relevance score for each topic; and

ranking the topics based on the sender and the receiver relevance scores.

6. The method according to claim 5, wherein computing the receiver relevance score comprises comparing a term frequency score for a topic to a corpus frequency score for the topic, wherein the corpus frequency score is based on a corpus of emails that have been at least one of sent or received by an intended recipient of the email message over a period of time.

7. The method according to claim 6, wherein computing the receiver relevance score further comprises identifying topics that match one or more characteristics of the intended recipient of the email message.

8. The method according to claim 7, wherein the one or more characteristics are derived from behavioral data of the intended recipient taken in connection with the corpus of emails.

9. The method according to claim 1, wherein the topics comprise at least one of key words or key phrases.

10. The method according to claim 1, further comprising assigning different subject lines to the email message for each respective recipient of a plurality of recipients of the email message.

11. The method according to claim 1, further comprising assigning a different subject line from the one or more of the subject lines assigned to the email message, the different subject line being visible to a sender of the email message and not visible to a recipient of the email message.

12. The method according to claim 1, wherein a plurality of the subject lines are assigned to the email message and are visible to an intended recipient in a configuration for displaying the plurality of the subject lines.