System and method for real-time identification of dissatisfaction data

Patent metadata

Field	Value
Publication number	US-2023197105-A1
Application number	US-202217992269-A
Country	US
Kind code	A1
Filing date	Nov 22, 2022
Priority date	Dec 22, 2021
Publication date	Jun 22, 2023
Grant date	—

Abstract

Various methods, apparatuses/systems, and media for automatic real-time identification of dissatisfaction data are disclosed. A receiver receives real-time stream of call transcript data generated during a call or a chat between a customer and an agent. A processor implements a machine learning model that includes predefined complaint data; applies the received call transcript data onto the machine learning model; compares, in response to applying, the call transcript data with predefined complaint data; generates a first similarity score, based on comparing, that identifies how similar the call transcript data is compared to the predefined complaint data; and automatically identifies the call transcript data as a first dissatisfaction data based on determining that the first similarity score is equal to or more than a predetermined threshold value.

First claim

What is claimed is:

1 . A method for automatic real-time identification of dissatisfaction data by utilizing one or more processors along with allocated memory, the method comprising:
establishing a communication link between an application and a cloud platform deployed on a cloud environment, the application including a service layer module, an interface layer module, and a user interface (UI) layer module; receiving, by the service layer module, real-time stream of call transcript data generated during a call or a chat between a customer and an agent; calling an application programming interface (API) to invoke the interface layer module to consume the real-time stream of the call transcript data as input data outputted from the service layer module; implementing a machine learning model that includes predefined complaint data; applying the call transcript data onto the machine learning model; comparing, in response to applying, the call transcript data with predefined complaint data; generating a first similarity score, based on comparing, that identifies how similar the call transcript data is compared to the predefined complaint data; and automatically identifying the call transcript data as a first dissatisfaction data based on determining that the first similarity score is equal to or more than a predetermined threshold value.

2 . The method according to claim 1 , wherein the call transcript data includes real-time stream of voice data received from a telephony device, and the method further comprising:
converting the voice data into text data; and applying the text data onto the machine learning model.

3 . The method according to claim 1 , wherein the call transcript data includes real-time stream of text data received from a chat platform, and the method further comprising:
generating text data from the call transcript data; and applying the text data onto the machine learning model.

4 . The method according to claim 1 , further comprising:
training the machine learning model with the first dissatisfaction data for automatic identification of a second dissatisfaction data and automatic identification of a second similarity score when a new call transcript data is received by the service layer module during a new call between the customer and the agent.

5 . The method according to claim 1 , in generating the call transcript data in real-time, the method further comprising:
concatenating all utterances during the call between the customer and the agent; converting the utterances into text data; and applying the text data onto the machine learning model.

6 . The method according to claim 1 , wherein the machine learning model includes one or more of the following models: a natural language processing (NLP) model and a long short term memory (LSTM) model.

7 . The method according to claim 6 , further comprising:
storing output from the machine learning model onto a database.

8 . The method according to claim 1 , further comprising:
automatically generating suggestions data based on the dissatisfaction data to resolve issues raised by the customer; and calling an API to invoke the UI interface layer module to display the suggestions data onto display screen of an agent computing device utilized by the agent.

9 . The method according to claim 8 , wherein the suggestions data includes one or more of the following data: complaint data; policy data; types of complaint data; sub-types of complaint data; and de-escalation script data.

10 . The method according to claim 9 , further comprising:
receiving agent feedback data on the complaint data from the agent's computing device; storing the agent feedback data onto a database; and consuming the agent feedback data by an analyst computing device utilized by an analyst.

11 . The method according to claim 10 , further comprising:
training the machine learning model by applying the agent feedback data onto the machine learning model; receiving, by the service layer module, a new call or a new chat between the customer and the agent; implementing the trained machine learning model; comparing, in response to implementing, the new call transcript data with the predefined complaint data; generating a second similarity score, based on comparing, that identifies how similar the new call transcript data is compared to the predefined complaint data; and automatically identifying the new call transcript data as a second dissatisfaction data based on determining that the second similarity score is equal to or more than the predetermined threshold value.

12 . The method according to claim 1 , wherein the predefined complaint data includes historical logs of complaint data within an organization along with complaint data accessed from public database of complaints, and the method further comprising:
training the machine learning model with the predefined complaint data.

13 . A system for automatic real-time identification of dissatisfaction data, the system comprising:
a processor; and a memory operatively connected to the processor via a communication interface, the memory storing computer readable instructions, when executed, causes the processor to: establish a communication link between an application and a cloud platform deployed on a cloud environment, the application including a service layer module, an interface layer module, and a user interface (UI) layer module; cause the service layer module to receive real-time stream of call transcript data generated during a call or a chat between a customer and an agent; call an application programming interface (API) to invoke the interface layer module to consume the real-time stream of the call transcript data as input data outputted from the service layer module; implement a machine learning model that includes predefined complaint data; apply the call transcript data onto the machine learning model; compare, in response to applying, the call transcript data with predefined complaint data; generate a first similarity score, based on comparing, that identifies how similar the call transcript data is compared to the predefined complaint data; and automatically identify the call transcript data as a first dissatisfaction data based on determining that the first similarity score is equal to or more than a predetermined threshold value.

14 . The system according to claim 13 , wherein the call transcript data includes real-time stream of voice data received from a telephony device, and the processor is further configured to:
convert the voice data into text data; and apply the text data onto the machine learning model.

15 . The system according to claim 13 , wherein the call transcript data includes real-time stream of text data received from a chat platform, and the processor is further configured to:
generate text data from the call transcript data; and apply the text data onto the machine learning model.

16 . The system according to claim 13 , wherein the processor is further configured to:
train the machine learning model with the first dissatisfaction data for automatic identification of a second dissatisfaction data and automatic identification of a second similarity score when a new call transcript data is received by the service layer module during a new call between the customer and the agent.

17 . The system according to claim 13 , in generating the call transcript data in real-time, the processor is further configured to:
concatenate all utterances during the call between the custom

Assignees

• Jpmorgan Chase Bank Na

Inventors

• Tanuku Siva
• Ray Matthew
• Brinda Rachel
• Wu Liping
• Wagner Tiffany
• Potluri Bhavya
• Sinha Khyati
• Kuyuk Serdar
• Joglekar Anand P
• Plantinga Peter
• Ni Kai

Classifications

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L15/04

  Segmentation; Word boundary detection · CPC title

• G06F40/20

  Natural language analysis (semantic analysis of natural language G06F40/30) · CPC title

• G10L15/30

  Distributed recognition, e.g. in client-server systems, for mobile phones or network applications · CPC title

• G10L25/63 Primary

  for estimating an emotional state · CPC title