The IQ application will be briefly unavailable on Sunday, March 31st, starting at 10:00am ET. Access will be restored as quickly as possible.
Browse Prior Art Database

Cognitive meeting assist mechanism

IP.com Disclosure Number: IPCOM000253595D
Publication Date: 2018-Apr-13
Document File: 3 page(s) / 22K

Publishing Venue

The IP.com Prior Art Database

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 44% of the total text.

Cognitive meeting assist mechanism

For many people, a typical work day requires attendance at multiple meetings. Meeting purposes include, but are not limited to, project management, status meetings, collaboration, etc. Having to attend many meetings often reduces the amount of productive work time a person has. In addition, many times, parts of a meeting are irrelevant to a person. If a person can ignore the irrelevant parts of meetings, then he/she will have more time to focus on tasks.

During strategy meetings, especially, discussions anchor around information collected prior to the meeting. This could include tedious work of reading and analyzing articles, reports, etc. that were published over a period. However, in the today’s dynamic world, where details change every minute and decisions need to be modified based on the most recent data, confining the meeting discussion to a past study is inadequate. Further, the time consumed in keeping a document alive is significant.

The strategist or decision maker(s) needs real-time updates before a meeting. Such a mechanism can not only save time, but also support accurate decision-making.

The novel contribution is a cognitive meeting assist mechanism for strategic decision making. The assist mechanism continuously reads substance and updates reports based on developments published on a day-to-day basis. In addition, this method helps a meeting attendee focus on relevant meeting topics and gives back to the attendee time consumed by irrelevant topics.

The novel system correlates the agenda item with the specific user based on content and the contextual situation, and then dynamically adds the content to the respective user’s repository for future referencing. The associated assist mechanism continuously reads substance and updates reports based on developments published on a day-to- day basis. It dynamically updates the user with the relevant details prior to the meeting based on the task assigned and contextual situation.

The core features needed for implementation of the solution are:

 Natural Language Processing (NLP) based keyword extraction. The mechanism uses this for extracting and storing relevant topics of conversation in a cloud database. The cloud database is integrated with the respective users who are engaged in that specific topic of interest.

 Mel-frequency cepstral coefficients (MFCCs) speech features extraction and feature matching. The MFCCs enable the mechanism to recognize current speakers and to whom the speech is referring.

To implement the solution, the system:

1. Builds a confidence level based on the above features

2. Creates tags with respect to the specific user and the associated voice (to establish relevance)

3. Establishes content and user relevance by:

A. Using the pattern history of the user

B. Monitoring the response of the user with respect to the specific parts of the conversation while the user is engaged in a meeting (e.g., a conference call)

4. Maps content...