System and method for data visualization on spatial computing device based on cascading machine learning approach

1 . A system, for displaying content on a display of a spatial computing device, the system comprising:

a memory configured to store a set of data records, each one of the set of data records including one or more data fields; and

a processor operably coupled to the memory, the processor configured to:

receive an audio signal associated with a call;

preprocess the audio signal to remove noise from the audio signal;

in response to removing the noise from the audio signal, extract one or more characteristics associated with the audio signal, wherein the one or more characteristics comprises spectral properties associated with the audio signal;

in response to extracting the one or more characteristics, identify a language of the call based on the spectral properties associated with the audio signal;

convert the audio signal into a communication transcript, wherein the communication transcript comprises one or more keywords and one or more character elements;

preprocess the communication transcript to remove the one or more character elements;

identify an intent of the communication transcript of the call based on the preprocessed communication transcript using a first natural language processing (NLP) machine learning model (MLM), wherein the first NLP MLM is trained on a first labeled dataset of historical communication transcripts, wherein first labels for the first labeled dataset represent intents of the historical communication transcripts;

identify the one or more keywords within the communication transcript using a second NLP MLM, the one or more keywords related to information stored in the set of data records;

select a set of target data records from the set of data records to associate with the communication transcript by:

determining for each data record from the set of data records, a probability score that each data record is associated with the communication transcript, wherein the probability score being determined by a data record prediction model; and

choosing the set of target data records being data records with corresponding probability scores having highest ranking, wherein a number of target data records within the set of target data records is a selected number;

generate a display report based on the selected set of target data records; and

render the display report on the display of the spatial computing device.

2 . The system of claim 1 , wherein identifying the language of the call is performed by using a language classifier, wherein the language classified is an MLM trained on a second labeled dataset of historical call data in different languages.

3 . The system of claim 2 , wherein second labels associated with the second labeled dataset represent languages used for the historical call data.

4 . The system of claim 3 , wherein the data record prediction model is configured to take input data comprising a data record and a text-string identifying a language and an intent, and output the probability score.

5 . The system of claim 4 , wherein the data record prediction model is trained using a third labeled dataset including:

the set of data records;

associated training input data, each one of the training input data containing information about a language and an intent; and

wherein third labels associated with the third labeled dataset represent a set of probability scores associated with the set of data records.

6 . The system of claim 1 , wherein the processor is further configured to:

receive the audio signal representing the call via a microphone associated with the spatial computing device.

7 . The system of claim 1 , wherein the generation of the display report comprises:

selecting from a plurality of display report templates corresponding to plurality of intents a target display report template corresponding to the identified intent, wherein each one of the plurality of display report templates includes report fields corresponding to at least some data fields of the set of target data records; and

populate the report fields of the target display report template with the at least some data fields of the set of target data records.

8 . The system of claim 1 , wherein the processor is further configured to:

select a template form, the template form including report fields; and

populate the report fields by automatically filling the template form with information obtained from the set of target data records.

9 . The system of claim 1 , wherein the set of target data records comprises tabulated data, and wherein the rendering of the display report comprises displaying the tabulated data.

10 . The system of claim 1 , wherein the call is between a caller and an agent, and wherein the communication transcript comprises a first labeled text corresponding to inquiries of the caller and a second labeled text corresponding to the communication of the agent.

11 . The system of claim 1 , wherein the processor is further configured to:

receive information about a type of the spatial computing device;

determine spatial integration instructions for rendering the display report based on the type; and

render the display report on the display consistent with the spatial integration instructions.

12 . The system of claim 1 , wherein the set of target data records is a first set of target data records, the selected number is the first selected number, and the display report is a first display report, the processor is further configured to:

receive feedback determining an accuracy of the first display report;

based on the feedback, when the first display report is determined to be inaccurate:

select a second set of target data records having corresponding probability scores ranking below the probability scores of the first set of target data records, wherein a number of target data records within the set of second target data records is a second selected number;

generate a second display report based on the second set target data records; and

render the second display report on the display of the spatial computing device.

13 . The system of claim 1 , wherein the processor is further configured to, based at least in part upon the identified intent:

request additional communication transcript data;

receive the additional communication transcript data; and

update the intent, based on the received additional communication transcript data.

14 . The system of claim 13 , wherein the processor is further configured to:

select another set of target data records from the set of data records to associate with the additional communication transcript data by:

determining for each data record from the set of data records, a probability score that each data record is associated with the additional communication transcript data; and

selecting the another set of target data records being data records with corresponding probability scores having highest ranking;

generate another display report based on the another set of target data records; and

render the another display report on the display of the spatial computing device.

15 . A method for displaying content on a display of a spatial computing device, the method comprising:

receiving an audio signal associated with a call;

preprocessing the audio signal to remove noise from the audio signal;

in response to removing the noise from the audio signal, extracting one or more characteristics associated with the audio signal, wherein the one or more characteristics comprises spectral properties associated with the audio signal;

in response to extracting the one or more characteristics, identifying a language of the call based on the spectral properties associated with the audio signal;

converting the audio signal into a communication transcript, wherein the communication transcript comprises one or more keywords and one or more character elements;

preprocessing the communication transcript to remove the one or more character elements;

identifying an intent of the communication transcript of the call based on the preprocessed communication transcript using a first natural language processing (NLP) machine learning model (MLM), wherein the first NLP MLM is trained on a first labeled dataset of historical communication transcripts, wherein first labels for the first labeled dataset represent intents of the historical communication transcripts;

identifying the one or more keywords within the communication transcript using a second NLP MLM, the one or more keywords related to information stored in a set of data records;

selecting a set of target data records from the set of data records to associate with the communication transcript by:

determining for each data record from the set of data records, a probability score that each data record is associated with the communication transcript, wherein the probability score being determined by a data record prediction model; and

choosing the set of target data records being data records with corresponding probability scores having highest ranking, wherein a number of target data records within the set of target data records is a selected number;

generating a display report based on the selected set of target data records; and

rendering the display report on the display of the spatial computing device.

16 . The method of claim 15 , further comprising:

receiving the audio signal representing the call via a microphone associated with the spatial computing device.

17 . The method of claim 15 , wherein the generation of the display report comprises:

selecting from a plurality of display report templates corresponding to plurality of intents a target display report template corresponding to the identified intent, wherein each one of the plurality of display report templates includes report fields corresponding to at least some data fields of the set of target data records; and

populating the report fields of the target display report template with the at least some data fields of the set of target data records.

18 . A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:

receive an audio signal associated with a call;

preprocess the audio signal to remove noise from the audio signal;

in response to removing the noise from the audio signal, extract one or more characteristics associated with the audio signal, wherein the one or more characteristics comprises spectral properties associated with the audio signal;

in response to extracting the one or more characteristics, identify a language of the call based on the spectral properties associated with the audio signal;

convert the audio signal into a communication transcript, wherein the communication transcript comprises one or more keywords and one or more character elements;

preprocess the communication transcript to remove the one or more character elements;

identify an intent of the communication transcript of the call based on the preprocessed communication transcript using a first natural language processing (NLP) machine learning model (MLM), wherein the first NLP MLM is trained on a first labeled dataset of historical communication transcripts, wherein first labels for the first labeled dataset represent intents of the historical communication transcripts;

identify the one or more keywords within the communication transcript using a second NLP MLM, the one or more keywords related to information stored in a set of data records;

select a set of target data records from the set of data records to associate with the communication transcript by:

determining for each data record from the set of data records, a probability score that each data record is associated with the communication transcript, wherein the probability score being determined by a data record prediction model; and

choosing the set of target data records being data records with corresponding probability scores having highest ranking, wherein a number of target data records within the set of target data records is a selected number;

generate a display report based on the selected set of target data records; and

render the display report on the display of a spatial computing device.

19 . The non-transitory computer-readable medium of claim 18 , wherein the instructions, when executed by the one or more processors, further cause the one or more processors to:

receive the audio signal representing the call via a microphone associated with the spatial computing device.

20 . The non-transitory computer-readable medium of claim 18 , wherein the generation of the display report comprises:

selecting from a plurality of display report templates corresponding to plurality of intents a target display report template corresponding to the identified intent, wherein each one of the plurality of display report templates includes report fields corresponding to at least some data fields of the set of target data records; and

populating the report fields of the target display report template with the at least some data fields of the set of target data records.