Conversation branching for more efficient resolution

Patent metadata

Field	Value
Publication number	US-2016253593-A9
Application number	US-201514661800-A
Country	US
Kind code	A9
Filing date	Mar 18, 2015
Priority date	Jun 3, 2014
Publication date	Sep 1, 2016
Grant date	—

Abstract

A method for conversation branching may include storing a plurality of messages communicated in an online conversation and generating a fingerprint for each message. The fingerprint for each message may be stored to a lookup table. The method may also include detecting a new message from one of the users and processing the new message to generate an artifact that correlates to the new message. The lookup table may be queried using the artifact to determine a similarity between the new message and each of the stored messages. One or more proposed reply messages to the new message may be generated for branching the online conversation to a new conversation point that corresponds to a selected one of the one or more proposed reply messages. The one or more proposed reply messages are based on one of the stored messages that has a closest similarity to the new message.

First claim

1 . A method for conversation branching, comprising:
storing, by a processor, a plurality of messages communicated in an online conversation between at least two users; generating, by the processor, a fingerprint for each message of the online conversation; storing, by the processor, the fingerprint for each message of the online conversation in a lookup table; detecting, by the processor, a new message from one of the users; processing, by the processor, the new message to generate an artifact that correlates to the new message; querying, by the processor, the lookup table using the artifact to determine a similarity between the new message and each of the stored messages; and generating, by the processor, one or more proposed reply messages to the new message for branching the online conversation to a new conversation point that corresponds to a selected one of the one or more proposed reply messages, the one or more proposed reply messages being based on one of the stored messages that has a closest similarity to the new message.

2 . The method of claim 1 , further comprising:
presenting the one or more proposed reply messages for selection by at least one user, wherein the new conversation point reduces or optimizes a number of additional new messages in the online conversation to reach a conclusion; and branching the online conversation to the new conversation point in response to the selected one of the one or more proposed reply messages being selected by the at least one user.

3 . The method of claim 1 , further comprising comparing the artifact to each fingerprint in the lookup table to identify a closest match between the new message and one of the stored messages.

4 . The method of claim 3 , wherein comparing the artifact to each fingerprint comprises comparing at least one of a set of keywords, a set of n-grams and parts-of-speech to each fingerprint.

5 . The method of claim 1 , wherein storing the fingerprint for each message of the online conversation in the lookup table comprises storing a pointer to the message associated with the fingerprint.

6 . The method of claim 1 , where storing the plurality of messages of the online conversation comprises storing each origin message paired with a reply message to the origin message.

7 . The method of claim 1 , wherein generating the fingerprint for each message comprises transforming each message into a series of n-grams.

8 . The method of claim 1 , wherein generating the fingerprint for each message comprises at least one of parsing each component of each message using a parse tree and using lemmatization.

9 . The method of claim 1 , wherein generating the fingerprint for each message comprises using a custom dictionary to identify keywords.

10 . The method of claim 1 , wherein generating the fingerprint for each message comprise pruning each message to eliminate unnecessary terms.

11 . The method of claim 1 , wherein a logical grouping of the fingerprints uniquely describes the online conversation.

12 . The method of claim 1 , further comprising receiving a set time window, wherein querying the lookup table comprises querying the lookup table during the set time window to determine a branch for the online conversation to the new conversation point.

13 . The method of claim 1 , wherein generating the fingerprint for each message comprises using natural language processing.

14 . The method of claim 1 , wherein each proposed reply message of the one or more proposed reply messages defines a different conversation path, and further comprising receiving a weighting for each different conversation path, the weighting corresponding to a prioritization for a particular outcome associated with each different conversation path.

15 - 20 . (canceled)

Assignees

• IBM

Inventors

• Bastide Paul R
• Bhogal Kulvir S
• Hoy Jeffrey R
• Silva Asima

Classifications

• G06Q10/107

  Computer-aided management of electronic mailing [e-mailing] · CPC title

• G06F40/205

  Parsing · CPC title

• G06F3/0482

  Interaction with lists of selectable items, e.g. menus · CPC title

• G06F16/9017

  using directory or table look-up (use of a directory or look-up table in file systems G06F16/13) · CPC title

• H04L51/02

  using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages · CPC title