Browse Prior Art Database

System and method for probabilistically determining unintended recipients using communication pattern recognition in an organization

IP.com Disclosure Number: IPCOM000234760D
Publication Date: 2014-Feb-03
Document File: 5 page(s) / 78K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a system and method for leveraging communication patterns in an organization to prevent unintended addressing. The idea rests on the premise that communication in any large organization isn't random but is influenced by organizational and geographical hierarchies, understanding which can help in probabilistically identifying an unintended recipient, whose name or communication address sounds, or is spelled similar to the intended recipient.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 31% of the total text.

Page 01 of 5

System and method for probabilistically determining unintended recipients using communication pattern recognition in an organization

In any large organization, it is common to have multiple users whose names and communication IDs sound or appear similar. And consequently, these users often receive communication intended for other users whose names are similar to theirs. Example- Vishal

Agarwal1/India/IBM and Vishal Agrawal1/India/IBM.

Serious consequences

Communicating with unintended recipients can have far more serious implications than mere interruptions, annoyances and productivity losses. One can inadvertently leak confidential or sensitive information, or perhaps wrongly assume that a certain action to his/her communication would to be taken by the recipient. For instance, such action could be responding to a time-bound sales opportunity, or preparing for an important client meeting. In cases where the unintended recipient does not notify the sender in time--away, traveling, or just doesn't care--these situations could lead to significant business losses.

And for someone whose name is frequent confused with others', inclusions into email threads or meeting invites can not only be frustrating but also leads to significant productivity loss as he/she must understand the communication context before responding to the sender accordingly. In cases where such thread is active and has large number of participants, it soon turns into a mess where the unintended recipient has to constantly request all participants to stop copying him/her.

Growing problem

The problem has grown recently due to two trends: workforce globalization and reliance on address completion algorithms. We're increasingly communicating with people from different geographies and cultures, many of whom we have never met and whose names are alien to us. So, we can't always tell the difference between an "Agarwal" and an "Agrawal", or an "Arti" and an "Aarti".

Current solutions

Modern collaboration clients like IBM Lotus Notes typically provide address completion functionality employing user's address book or an organization wide address database such as LDAP, in conjunction with user's communication history. While this approach works most of the time, a small typo in recipient's name or part thereof can cause the address completion algorithm fail or worse, present a different result. And if the user isn't careful, he/she would not only start communication with an unintended recipient but also add the unintended recipient in his/her own address book and communication history, thereby increasing the likelihood of communicating with the same unintended recipient again.

Present address completion algorithms are also ill-equipped to predict with reasonable accuracy

1


Page 02 of 5

the intended recipient, especially one the sender has never communicated with directly . Usually, one simply types in the recipient's name (as he/she feels it should be spelled) and if there's n...