US 12,174,873 B2
Dynamic prediction of agenda item coverage in a communication session
Ling Tsou, Lawndale, CA (US); Tong Wang, Lexington, MA (US); and Yun Zhang, Pittsburgh, PA (US)
Assigned to Zoom Video Communications, Inc., San Jose, CA (US)
Filed by Zoom Video Communications, Inc., San Jose, CA (US)
Filed on Oct. 31, 2022, as Appl. No. 17/978,072.
Prior Publication US 2024/0143643 A1, May 2, 2024
Int. Cl. G06F 16/35 (2019.01); G06F 16/383 (2019.01); G06F 40/284 (2020.01); H04L 12/18 (2006.01)
CPC G06F 16/353 (2019.01) [G06F 16/383 (2019.01); G06F 40/284 (2020.01); H04L 12/1831 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method, comprising:
connecting to a communication session with a plurality of participants;
receiving one or more agenda items for the communication session;
receiving a plurality of utterances from participants produced during the communication session;
classifying each agenda item as a long item or a short item, a short item being covered within a specified sentence threshold;
for agenda items classified as long items:
extracting one or more topics from the utterances via a sliding window method, wherein the sliding window method comprises:
shifting a window over the utterances one word at a time with a pre-specified window size to generate two blocks of utterances per each shift of the window;
at each shift of the window, comparing the two blocks of the utterances to determine whether the blocks are semantically similar; and
defining a boundary between two topics when two blocks of utterances are semantically different, and
using a topic detection model to determine whether a topic related to each agenda item classified as a long item has been covered;
for agenda items classified as short items:
applying one or more matching methods to determine whether one or more of the utterances within the specified sentence threshold cover each agenda item classified as a short item; and
transmitting, to one or more client devices, a status of the agenda items for the communication session.