EMAIL ANALYTICS

• US 20150269242A1
• Filed: 10/21/2014
• Published: 09/24/2015
• Est. Priority Date: 03/20/2014
• Status: Active Grant

• Current Assignee: TATA Consultancy Services Limited (Tata Sons Pvt Ltd.)
• Original Assignee: TATA Consultancy Services Limited (Tata Sons Pvt Ltd.)
• Inventors: DEY, Lipika, SHROFF, Gautam, SINGH, Somya
• Granted Patent: US 9,710,539 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: H04L 51/00 User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
• CPC Class Codes: H04L 51/216 Handling conversation history, e.g. grouping of messages in sessions or threads
• CPC Class Codes: H04L 51/234 for tracking messages

Abstract

A method for performing email analytics is described. The method includes extracting emails from the configured email repository. The emails are then grouped into mail groups based on identification of content similarity of the emails. A network graph is then constructed for each of the mail group to identify an association of emails in the mail group based on header-level analysis of emails. Thereafter, email analytics is performed on the mail groups by clustering the mail groups into mail clusters based on temporal progression of emails in the mail groups. Key phrases are then determined based on a content analysis of emails in the mail groups in the mail clusters. The key phrases are then associated with the network graphs of the mail groups.

Claims

1. A computer implemented method for performing email analytics on a configured email repository, the computer implemented method comprising:

grouping, by a processor, emails in the configured email repository, into a plurality of mail groups based on identifying similarity of content in the emails, wherein each of the plurality mail groups is associated with a timeline;

determining, by the processor, a network graph for each of the plurality of mail groups based on a header information of emails within each of the plurality of mail groups, wherein the network graph indicates an association of the emails in the plurality of mail groups; and

performing, by the processor, email analytics on the plurality of mail groups, wherein the performing includes;

clustering, by the processor, the plurality of mail groups into a plurality of mail clusters based on a temporal analysis of the emails within each of the email groups, wherein the temporal analysis identifies a temporal progression of the emails over the timeline of each of the plurality of mail groups;

determining, by the processor, key phrases from content of emails in the plurality of mail groups in each of the plurality of mail clusters; and

associating, by the processor, the key phrases with the network graph of each of plurality of mail groups, wherein the key phrases are indicative of type of activity and state of activity in each of the plurality of mail groups.

2. The computer implemented method as claimed in claim 1, wherein the key phrases are associated with the network graph by the processor using tag-clouds identifying the key phrases in accordance with a predetermined importance.

3. The computer implemented method as claimed in claim 1, wherein the determination of the key phrases, by the processor, is based on a content analysis of the emails within the plurality of mail groups in each of the mail clusters and wherein the key phrases include named entities and user interaction information.

4. The computer implemented method as claimed in claim 1 further comprising extracting, by the processor, emails from the configured email repository for grouping the emails into the plurality of mail groups.

5. The computer implemented method as claimed in claim 1, wherein the determination of the network graph for each of the plurality of mail groups by the processor is further based on analysis of the header information of emails within each of the plurality of mail groups.

6. The computer implemented method as claimed in claim 1, wherein the network graph identifies the association of emails in the mail group based on an order of occurrence of the emails in the mail group.

7. The computer implemented method as claimed in claim 1, wherein the grouping of email into the plurality of mail groups by the processor further comprises identifying overlapping content in the emails based on email level similarity in the emails and trail mail level similarity in trail mails of the emails in each of the plurality of mail groups.

8. An email analytics system for performing email analytics on a configured email repository, the email analytics system comprising:

a processor;

a mail reconstruction module coupled to the processor to;

group emails in the configured email repository into a plurality of mail groups based on identifying similarity of content in the emails, wherein each of the plurality mail groups is associated with a timeline; and

determine a network graph for each of the plurality of mail groups based on a header information of emails within each of the plurality of mail groups, wherein the network graph indicates an association of emails in the plurality of mail groups; and

an email analytics module coupled to the processor to perform email analytics on the plurality of mail groups, wherein the email analytics module performs the following;

cluster the plurality of mail groups into a plurality of mail clusters based on the timeline of each of the plurality of mail groups;

determine key phrases from content of emails in the plurality of mail groups in each of the plurality of mail clusters; and

associate the key phrases with the network graph of each of plurality of mail groups, wherein the key phrases are indicative of type of activity and state of activity of each of the plurality of mail groups.

9. The email analytics system as claimed in claim 8, wherein the email analytics module further comprises a temporal analysis module to cluster the plurality of mail groups into the plurality of mail clusters based on a temporal progression of the emails over the timeline of each of the plurality of mail groups.

10. The email analytics system as claimed in claim 9, wherein the temporal analysis module determines the temporal progression of the emails over the timeline of each of the plurality of mail groups by computing activity density distribution of each of the plurality of mail groups.

11. The email analytics system as claimed in claim 8, wherein the email analytics module further comprises a content analysis module to:

determine key phrases from content of emails in the plurality of mail groups in each of the plurality of mail clusters based on a content analysis of each of the mail clusters, wherein the key phrases include named entities and user interaction information; and

associate key phrases with the network graph using tag-clouds, wherein the tag-clouds indicate the key phrases in accordance with an associated importance.

12. The email analytics system as claimed in claim 8, wherein an email extraction module is coupled to the processor to extract emails from the configured email repository for performing email analytics.

13. The email analytics system as claimed in claim 8, wherein the mail reconstruction module includes a network reconstruction module for determining the network graph for each of the plurality of mail groups to identify an order of occurrence of the emails in each of the plurality of mail groups.

14. The email analytics system as claimed in claim 8, wherein the mail reconstruction module further comprises a duplicate detection module to identify similarity of content similarity based on overlapping content in the emails, the identification including identifying email level similarity in the emails and trail mail level similarity in trail mails of the emails in the plurality of mail groups.

15. A non-transitory computer-readable medium comprising instructions executable by a email analytics system to:

group emails in a configured email repository, into a plurality of mail groups based on identifying similarity of content in the emails, wherein each of the plurality mail groups is associated with a timeline;

determine a network graph for each of the plurality of mail groups based on a header information of emails within each of the plurality of mail groups, wherein the network graph indicates an association of emails in the mail group; and

perform email analytics on the plurality of mail groups, wherein the performing includes;

cluster the plurality of mail groups into a plurality of mail clusters based on a temporal analysis of the emails within each of the email groups, wherein the temporal analysis identifies a temporal progression of the emails over the timeline of each of the plurality of mail groups;

determine key phrases from content of emails in the plurality of mail groups in each of the plurality of mail clusters based on a content analysis of the emails within the plurality of mail groups in each of the mail clusters, wherein the key phrases include named entities and user interaction information; and

associate the key phrases with the network graph of each of plurality of mail groups, wherein the key phrases are indicative of type of activity and state of activity in each of the plurality of mail groups.

16. The non-transitory computer-readable medium as claimed in claim 15, wherein the non-transitory computer-readable medium further comprises instructions executable by the email analytics system to cluster the plurality of mail groups into the plurality of mail clusters based on a temporal progression of the emails over the timeline of each of the plurality of mail groups.

17. The non-transitory computer-readable medium as claimed in claim 16, wherein the non-transitory computer-readable medium further comprises instructions executable by the email analytics system to determine the temporal progression of the emails over the timeline of each of the plurality of mail groups by computing activity density distribution of each of the plurality of mail groups.

18. The non-transitory computer-readable medium as claimed in claim 15, wherein the non-transitory computer-readable medium further comprises instructions executable by the email analytics system to identify similarity of content similarity based on overlapping content in the emails, the identification including identifying email level similarity in the emails and trail mail level similarity in trail mails of the emails in the plurality of mail groups.