Devices and methods involving time series signal assessment
Aspects of the present disclosure are directed to computer-implemented methods, apparatuses and systems involving the processing of time series signals for assessing trend characteristics related thereto. In a particular aspect, a time series of signals generated at respective endpoint devices participating in data communications in a data-communications system are monitored. Categorical variables indicative of a characteristic trend of the time series of signals are assessed, and for one or more of the endpoints that communicates the time series of signals, a subset of the time series of signals that effect a change in the characteristic trend of the signals is identified. An output characterizing the change and the subset of the time series of signals is generated and transmitted.
1 . A computer-implemented method comprising:
hosting a meeting in a data-communications system during which a plurality of participants communicate with each other via a respective ones of a plurality of endpoint devices, such that the plurality of endpoint devices generates a plurality of signals respectively indicative of respective communications made by the plurality of participants during the meeting;
monitoring the plurality of signals;
deriving a plurality of time series of user-level categorical variable data respectively from the monitored plurality of signals, the plurality of time series of user-level categorical variable data respectively indicating sentiments of respective ones of the plurality of participants during the meeting;
determining multiple slope segments of each of the plurality of time series of user-level categorical variable data, the respective multiple slope segments being indicative of a respective user-level trend in the respective sentiment of each of the respective plurality of participants during the meeting;
analyzing the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data to identify one of the plurality of time series of user-level categorical variable data having a change in the user-level trend in the sentiment of a respective one of the plurality of participants during the meeting; and
generating and transmitting an output to an endpoint device identifying the change in the user-level trend in the sentiment of the respective participant during the meeting.
2 . The method of claim 1 , wherein generating and transmitting the output to the endpoint device comprises presenting a graphical indication of the change in the user-level trend for display on the endpoint device.
3 . The method of claim 1 , wherein the meeting is a videoconference, and wherein the plurality of signals are video signals.
4 . The method of claim 1 , further comprising controlling subsequent communications between the plurality of participants in response to generating and transmitting an output to the endpoint device.
5 . The method of claim 1 , wherein the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data are indicative of a rate of change in the user-level trend in the sentiment of each of the plurality of participants during the meeting, wherein the identified time series of user-level categorical variable data has a rate of change in the user-level trend greater than a threshold, and wherein
the output to the endpoint device identifies the rate of change in the user-level trend in the sentiment of the respective participant.
6 . The method of claim 1 , wherein the output identifies the endpoint device that generated the time series of the respective one of the plurality of signals from which the identified time series of time series of user-level categorical variable data was derived.
7 . The method of claim 1 , wherein the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data is analyzed to identify the one of the plurality of time series of user-level categorical variable data having a change in the user-level trend in the sentiment of the respective participant during the meeting greater than a change in user-level trends in the sentiment of others of the plurality of participants during the meeting.
8 . The method of claim 1 ,
further comprising transforming a summation of the slope segments of each of the plurality of time series of user-level categorical variable data into a time series of summed slope values, wherein the output to the endpoint device identifies the time series of summed slope values.
9 . The method of claim 1 , wherein each of the determined multiple slope segments of each of the plurality of time series of user-level categorical data indicates either a positively increasing sentiment or a negatively increasing sentiment of each of the respective plurality of participants during the meeting.
10 . The method of claim 1 , further comprising:
deriving a time series of meeting-level categorical variable data from the monitored plurality of signals, the time series of meeting-level categorical variable data indicating a sentiment of the meeting;
wherein the plurality of time series of user-level categorical variable data is respectively a plurality of time series of user-level categorical variable data respectively correlated to the time series of meeting-level categorical variable data;
wherein the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data is indicative of a correlated user-level trend in the sentiment of each of the respective plurality of participants during the meeting;
wherein the identified one of the plurality of time series of correlated user-level categorical variable data has a change in the correlated user-level trend in the sentiment of the respective participant during the meeting; and
wherein the output to the endpoint device identifies the change in the correlated user-level trend in the sentiment of the respective participant during the meeting.
11 . The method of claim 10 , further comprising:
transforming a summation of the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data into a time series of summed slope values; and
identifying locations of transitions between a sign of the time series of correlated summed slope values;
wherein the output to the endpoint device further identifies the locations of the transitions.
12 . A data communications system, comprising:
server circuitry configured for hosting a meeting in a data-communications system during which a plurality of participants communicate with each other via a respective ones of a plurality of endpoint devices, such that the plurality of endpoint devices generates a plurality of signals respectively indicative of respective communications made by the plurality of participants during the meeting; and
analytics circuitry configured for:
monitoring the plurality of signals;
deriving a plurality of time series of user-level categorical variable data respectively from the monitored plurality of signals, the plurality of time series of user-level categorical variable data respectively indicating sentiments of respective ones of the plurality of participants during the meeting;
determining multiple slope segments of each of the plurality of time series of user-level categorical variable data, the respective multiple slope segments being indicative of a respective user-level trend in the respective sentiment of each of the respective plurality of participants during the meeting;
analyzing the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data to identify one of the plurality of time series of user-level categorical variable data having a change in the user-level trend in the sentiment of a respective one of the plurality of participants during the meeting; and
generating and transmitting an output to an endpoint device identifying the change in the user-level trend in the sentiment of the respective participant during the meeting.
13 . The data communications system of claim 12 , wherein the analytics circuitry is further configured for transforming a summation of the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data into a time series of summed slope values, wherein the output to the endpoint device identifies the time series of summed slope values.
14 . The data communications system of claim 12 , wherein each of the determined multiple slope segments of each of the plurality of time series of user-level categorical data indicates either a positively increasing sentiment or a negatively increasing sentiment of each of the respective plurality of participants during the meeting.
15 . The data communications system of claim 12 ,
wherein the analytics circuitry is further configured for deriving a time series of meeting-level categorical variable data from the monitored plurality of signals, the time series of meeting-level categorical variable data indicating a sentiment of the meeting;
wherein the plurality of time series of user-level categorical variable data is a plurality of time series of user-level categorical variable data correlated to the time series of meeting-level categorical variable data;
wherein the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data is indicative of a correlated user-level trend in the sentiment of each of the respective plurality of participants during the meeting;
wherein the identified one of the plurality of time series of correlated user-level categorical variable data has a change in the correlated user-level trend in the sentiment of the respective participant during the meeting; and
wherein the output to the endpoint device identifies the change in the correlated user-level trend in the sentiment of the respective participant during the meeting.
16 . The data communications system of claim 12 , wherein the analytics circuitry is further configured for:
transforming a summation of the determined multiple slope segments of each of the plurality of time series of user-level categorical variable data into a time series of summed slope values; and
identifying locations of transitions between a sign of the time series of correlated summed slope values;
wherein the output to the endpoint device further identifies the locations of the transitions.
17 . A non-transitory computer-readable storage medium comprising instructions stored thereon, which when executed by one or more processors, cause the one or more processors to perform operations for managing a meeting in a data-communications system during which a participant communicates with other participants via an endpoint device, such that the endpoint device generates a signal indicative of communications made by the participant during the meeting, the operations comprising:
monitoring the signal;
deriving a time series of user-level categorical variable data from the monitored signal, the time series of user-level categorical variable data indicating a sentiment of the participant during the meeting;
determining multiple slope segments of the time series of user-level categorical variable data, the multiple slope segments being indicative of a user-level trend in the sentiment of each of the respective plurality of participants during the meeting;
analyzing the determined multiple slope segments of the time series of user-level categorical variable data to identify a change in the user-level trend in the sentiment of the participant during the meeting; and
generating and transmitting an output to an endpoint device identifying the change in the user-level trend in the sentiment of the participant during the meeting.
18 . The non-transitory computer readable storage medium of claim 17 , further comprising controlling subsequent communications from the participant in response to generating and transmitting an output to the endpoint device.
19 . The non-transitory computer-readable storage medium of claim 17 , wherein each of the determined multiple slope segments of the time series of user-level categorical data indicates either a positively increasing sentiment or a negatively increasing sentiment of the participant during the meeting.
20 . The non-transitory computer-readable storage medium of claim 17 , wherein the operations further comprise:
monitoring a plurality of signals respectively generated by a plurality of endpoint devices during the meeting, the plurality of signals indicative of respective communications made by a plurality of participants during the meeting;
deriving a time series of meeting-level categorical variable data from the monitored plurality of signals, the time series of meeting-level categorical variable data indicating a sentiment of the meeting;
wherein the time series of user-level categorical variable data is a time series of user-level categorical variable data correlated to the time series of meeting-level categorical variable data;
wherein the determined multiple slope segments of the time series of user-level categorical variable data is indicative of a correlated user-level trend in the sentiment of the participant during the meeting;
wherein the identified change in the user-level trend in the sentiment of the participant during the meeting is a change in the correlated user-level trend in the sentiment of the participant during the meeting; and
wherein the output to the endpoint device identifies the change in the correlated user-level trend in the sentiment of the participant during the meeting.