Browse Prior Art Database

Language Processing System for Multiple Speakers Disclosure Number: IPCOM000248642D
Publication Date: 2016-Dec-22
Document File: 1 page(s) / 19K

Publishing Venue

The Prior Art Database


Disclosed is a system that uses the transcription of a meeting along with cognitively derived data to create and execute action items given during the meeting.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.


Language Processing System for Multiple Speakers

During meetings, it is common for facilitators and attendees to create many action items. Normally, an attendee of the meeting records notes or action items. To make this task more efficient and cost-effective, an automated method is needed.

The novel contribution is a system that can create and execute cognitive actions based on the transcription of a meeting along with cognitively derived data such as the identified speaker and the person to whom the speaker gave direction.

To implement this system in a preferred embodiment: 1. A voice transcription system transcribes or otherwise interprets the spoken

contents of a meeting 2. The system uses cognitive analysis to determine context of the interpretation of

the meeting including: A. Speaker B. Receiver C. Speakers in the organization that are not present D. References to previous or future subjects

3. The system continuously checks for actionable commands spoken by the participants, such as "schedule a meeting for x", "remind me to do x", "invite x to the next meeting", "email x about y", "send an email to x saying y", etc.

4. If the system determines it has enough information to immediately execute an action, it does execute the action. Otherwise, the system creates a list for each participant with outstanding action items, and then presents the list to the participant after the meeting.