Browse Prior Art Database

CONVERSATION ROOM TOPIC ADHERANCE

IP.com Disclosure Number: IPCOM000242967D
Publication Date: 2015-Sep-03
Document File: 5 page(s) / 682K

Publishing Venue

The IP.com Prior Art Database

Related People

John Lynch: AUTHOR [+3]

Abstract

A system is presented that monitors and measures room topic deviation informing room owners and users of topic contextual adherence. Advantages of this system include maintaining context for persistent chat sessions, and informing owners and users that they are deviating from topic. This deviation may be acceptable or even desirable but the indicator is still valuable. The system can also suggest alternative room names based where discussion topic deviates from the original topic.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 5

CONVERSATION ROOM TOPIC ADHERANCE

AUTHORS:

John Lynch

 Keith Griffin
Katarina Lukacsy

CISCO SYSTEMS, INC.

ABSTRACT

    A system is presented that monitors and measures room topic deviation informing room owners and users of topic contextual adherence. Advantages of this system include maintaining context for persistent chat sessions, and informing owners and users that they are deviating from topic. This deviation may be acceptable or even desirable but the indicator is still valuable. The system can also suggest alternative room names based where discussion topic deviates from the original topic.

DETAILED DESCRIPTION

     Conversation rooms and persistent chat groups can often deviate from their original topic or purpose. For example, a chat room is started to discuss Windows® clients but evolves to a room discussing multiple client platforms and operating systems.

    Presented herein is a system that monitors and measures room topic adherence with a view towards maintaining accurate context for room owners and users.

The system consists of:


- Conversation room / Persistent Chat Room.

- A semantic monitoring agent.

- An adherence analysis agent.

- A presentation module to display adherence results.

Copyright 2015 Cisco Systems, Inc.

1


Page 02 of 5

FIG. 1 below shows a block diagram of the topic adherence system.

FIG. 1

    The conversation room can be thought of as a multi-party chat room such as Persistent Chat Room. Rooms in the system will have a semantic monitoring agent that will extract meaningful topics from the room using natural language processing (NLP).

    The meaningful topic is periodically compared to the original room name and description by the adherence analysis agent and is scored based on relevance. The result is plotted by the presentation module and displayed in the room user interface (UI) to the room owner and/or room participants. An example of the output from the adherence analysis agent is shown below in FIG. 2.

Copyright 2015 Cisco Systems, Inc.

2


Page 03 of 5

FIG. 2

    FIG. 2 shows how the deviation from the original room topic is determined using NLP and topic hierarchies. FIG. 2 uses the example given above. Over a period of time the Windows Client room members start to have discussions about Mac and Linux support while discussions about Windows clients drop off in parallel. Then the adherence agent would show a deviation from the original room topic and suggest that the room could be renamed "Desktop Clients". In an extended example, if the room members start discussing iOS and Android as examples of a Mobile OS, the suggestion could become "Client Operating Systems". This is functionally achieved using the NLP systems hierarchy of terminology which is common in such systems.

    FIG....