Browse Prior Art Database

System and method to cognitively organize meetings from multi-person chats

IP.com Disclosure Number: IPCOM000250469D
Publication Date: 2017-Jul-24
Document File: 5 page(s) / 212K

Publishing Venue

The IP.com Prior Art Database

Abstract

Chatrooms, such as Sametime Chatrooms, Slack.com provide a means for real time textual communication between team members. Sometimes it is determined that an actual meeting, virtual or face to face should occur.

Currently there is no easy way to quickly organize the meeting based on the textual chat and attendees. In addition, people who are not present in the chat (due to illness, vacation, lunch, other meetings) can struggle to follow long rambling multi-person chats.

It is proposed that the chatroom has a virtual assistant, hereby referred to as @Assist, who is part of the chat. @Assist is monitoring the chat using the cognitive APIs. Other (real people) chat members can address @Assist for assistance in setting up meetings and in addition @Assist will recognize when a meeting is being proposed and offer to set it up. @Assist will create a draft meeting invite that includes a background and an agenda that has been garnered from the recent chat. As @Assist has the ability to learn this may include information and background from past chats.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 32% of the total text.

TITLE: System and method to cognitively organize meetings from multi-person chats ABSTRACT:

Chatrooms, such as Sametime Chatrooms, Slack.com provide a means for real time textual communication between team members. Sometimes it is determined that an actual meeting, virtual or face to face should occur. Currently there is no easy way to quickly organize the meeting based on the textual chat and attendees. In addition, people who are not present in the chat (due to illness, vacation, lunch, other meetings) can struggle to follow long rambling multi-person chats. It is proposed that the chatroom has a virtual assistant, hereby referred to as @Assist, who is part of the chat. @Assist is monitoring the chat using the cognitive APIs. Other (real people) chat members can address @Assist for assistance in setting up meetings and in addition @Assist will recognize when a meeting is being proposed and offer to set it up. @Assist will create a draft meeting invite that includes a background and an agenda that has been garnered from the recent chat. As @Assist has the ability to learn this may include information and background from past chats. The idea should help teams, in particular geographically dispersed teams, work more efficiently. Meetings will become quicker and easier to setup with the help of the automated invite. Members who missed or failed the follow the chat thread will get a summary of the chat when subsequently invited to a meeting. Members can create and organize meetings more easily. The method comprises the following steps:

1. Define the chatroom’s purpose – why are these group of people together 1. Expert people feed all relevant documentation pertaining to this purpose to @Assist

1. @Assist extracts and creates a topic glossary (e.g. Networking, backups, oslc). Step is repeated as documents are updated. @Assist may for example subscribe to an RSS feed on wikis.

2. Expert people review and refine the topic glossary 2. Program @Assist to react to the term or synomyns of «meeting» 3. @Assist subscribes to or listens to the chatroom. This reacts to the users in real-time and also

can maintain datastores of the messages for learning. E.g. Pub-sub with some sort of historian. 4. @Assist starts to build a dataset relating the topics in the glossary identified in step 1 with the

users of the chatroom. 5. User asks for a meeting, either directly by clicking on a button or indirectly by using the word

meeting in a comment 1. @Assist reacts. For example, it may light up or change color or appear to be typing. 2. @Assist sets up a meeting invite with a sample agenda and invitee list

1. The agenda is based on the preceding chat. Initally @Assist will take from where there was time gap (for example 30 minutes).

2. The user adjusts the agenda. @Assist stores the adjustment and seeks to learn to make better first cut agendas. Based on adjustments to the agenda, the topic glossary may also be updated with new topics

3. @Assist identifies the topics...