Predictive analytic systems and methods

• US 10,262,268 B2
• Filed: 10/04/2013
• Issued: 04/16/2019
• Est. Priority Date: 10/04/2013
• Status: Active Grant

• Current Assignee: Mattersight Corporation (Nice Ltd)
• Original Assignee: Mattersight Corporation (Nice Ltd)
• Inventors: Danson, Christopher, McIngvale, Brittney Lynn, Stephan, Alain, Stadnik, Tomasz, Desikan, Samantha Shruti, Warford, Roger
• Primary Examiner(s): Cassity, Robert A
• Assistant Examiner(s): Zidanic, Michael
• Application Number:
  US14/045,968
• Publication Number:
  US 20150100528A1
• Time in Patent Office: 2,020 Days
• Field of Search: 706 21
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

The methods, apparatus, and systems described herein facilitate decision-making by providing predictions of outcomes and behaviors. The methods include receiving a communication between an agent and a prospect, analyzing density of keywords in a text version of the communication to determine the type of communication and amount of value time, determining if the communication is a first meaningful contact based on the type of communication and amount of value time, and predicting a likelihood of a prospect'"'"'s action based on the determination.

Claims

1. A system for routing outbound communications to a prospect based on a prior communication, comprising:

an Internet Protocol Private Branch Exchange (IP/PBX) circuit; and

a node comprising a processor and a non-transitory computer readable medium operably coupled thereto, the non-transitory computer readable medium comprising a plurality of instructions stored in association therewith that are accessible to, and executable by, the processor, where the plurality of instructions comprises;

instructions, that when executed, receive a prediction request from a user interface application on a client device;

instructions, that when executed, record a communication between an agent and a prospect on the client device and a telecommunication device of the prospect;

instructions, that when executed, transcribe the recorded communication from audio to text when the recorded communication is not in text form;

instructions, that when executed, determine which segments of the text of the recorded communication originated with the agent, and which segments of the text of the recorded communication originated with the prospect;

instructions, that when executed, apply linguistic analysis to the text of the recorded communication to determine a type of communication, a density of keywords relating to an organization, business, service or product, or any combination thereof, and a relationship of these keywords to each other during the recorded communication, wherein the keywords are selected from a library and are related to the type of communication;

instructions, that when executed, apply a linguistic-based psychological behavioral model to the text of the recorded communication to determine a personality type of the agent and the prospect based on identifying keywords relevant to the applied linguistic-based psychological behavioral model;

instructions, that when executed, separate agent voice data from the text of the recorded communication;

instructions, that when executed, calculate a weight of keywords identified in a first segment of the agent voice data relative to other keywords identified in the first segment or a second segment in close proximity to the first segment;

instructions, that when executed, calculate an amount of time in the agent voice data that the agent used the keywords relating to the organization, business, service, or product, or any combination thereof;

instructions, that when executed, determine whether the calculated amount of time exceeds a threshold time, wherein the threshold time is based on the type of communication;

instructions, that when executed, generate a predictive output for the user interface application on the client device based on the determined personality types of the agent and the prospect, the type of communication, and the calculated weighting of the identified keywords, when the calculated amount of time exceeded the threshold time;

instructions, that when executed, access a database storing agent history with a set of prospects having the determined personality type of the prospect;

instructions, that when executed, select an agent based on the agent history with the set of prospects having the determined personality type of the prospect;

instructions, that when executed, connect, via the IP/PBX circuit, the prospect on a subsequent communication to a workstation of the selected agent.

2. The system of claim 1, further comprising instructions, that when executed, determine at least an initial meaningful contact made between the agent and the prospect.

3. The system of claim 2, wherein the meaningful contact is used as an additional input to generate the predictive output.

4. The system of claim 1, further comprising instructions, that when executed, measure a level of engagement between the agent and the prospect.

5. The system of claim 4, further comprising instructions, that when executed, use the level of engagement as an additional input to determine if the communication is a meaningful contact.

6. The system of claim 1, further comprising instructions, that when executed, use demographic information, behavioral analytics data, prior communication information, or combinations thereof as additional inputs to generate the predictive output.

7. The system of claim 1, wherein the prospect is a prospective student.

8. The system of claim 7, further comprising instructions, that when executed, collect information regarding student attendance, participation, grades, or a combination thereof, and update the likelihood of the prospective student'"'"'s action.

9. The system of claim 1, wherein the prospect is a prospective buyer.

10. The system of claim 9, further comprising instructions, that when executed, assign the prospective buyer to the selected agent.

11. The system of claim 10, wherein the agent history comprises history of an agent with the prospective buyer'"'"'s personality type.

12. A method for routing outbound communications to a prospect based on a prior communication, which comprises:

receiving, by one or more processors, a prediction request from a user interface application on a client device;

recording, by the one or more processors, a communication between an agent and a prospect on the client device and a telecommunication device of the prospect, the communication having a density of keywords;

transcribing the recorded communication from audio to text when the recorded communication is not in text form;

determining which segments of the text of the recorded communication originated with the agent, and which segments of the text of the recorded communication originated with the prospect;

analyzing, by the one or more processors, the density of keywords relating to an organization, business, service or product, or any combination thereof, and a relationship of these keywords to each other, in the text of the recorded communication, wherein the keywords are selected from a library and are related to a type of communication;

applying, by the one or more processors, a linguistic-based psychological behavioral model to the text of the recorded communication to determine a personality type of the agent and the prospect based on identifying keywords relevant to the applied linguistic-based psychological behavioral model;

separating, by the one or more processors, agent voice data from the text of the recorded communication;

calculating, by the one or more processors, a weight of keywords identified in a first segment of the agent voice data relative to other keywords identified in the first segment or a second segment in close proximity to the first segment;

calculating, by the one or more processors, an amount of time in the agent voice data that the agent used the keywords relating to the organization, business, service or product, or any combination thereof;

determining, by the one or more processors, whether the calculated amount of time exceeds a threshold time, wherein the threshold time is based on the type of communication;

determining, by the one or more processors, if the recorded communication is a first meaningful contact based on the type of communication and the density of keywords during the recorded communication;

generating, by the one or more processors, a predictive output for the user interface application on the client device based on the determination of the personality types for the agent and the prospect, the type of communication, and the calculated weighting of the identified keywords, when the calculated amount of time exceeded the threshold time and when the recorded communication includes the first meaningful contact;

accessing, by the one or more processors, a database storing agent history with a set of prospects having the determined personality type of the prospect;

selecting, by the one or more processors, an agent based on the agent history with the set of prospects having the determined personality type of the prospect;

connecting, by an Internet Protocol Private Branch Exchange (IP/PBX) circuit, the prospect on a subsequent communication to a workstation of the selected agent.

13. The method of claim 12, which further comprises measuring a level of engagement between the agent and the prospect.

14. The method of claim 13, wherein the determining further comprises using the level of engagement to determine if the communication is the first meaningful contact.

15. The method of claim 12, which further comprises using demographic information, behavioral analytics data, prior communication information, or a combination thereof as an additional input to generate the predictive output.

16. The method of claim 12, wherein the calculated amount of time the agent voice data is associated with the density of keywords and is measured by an associated length of time in the agent voice data during which the agent is promoting a value of the service, product, or combination thereof to the prospect.

17. The method of claim 12, wherein the prospect is a prospective student.

18. The method of claim 17, wherein the contact type comprises an inquiry, enrollment, encouragement, or advice.

19. The method of claim 17, wherein the prospective student'"'"'s action comprises applying to a school, visiting a school, enrolling in a program, attending a class, graduating from a class or a program, finishing at the top of a class, or a combination thereof.

20. The method of claim 17, which further comprises collecting information regarding student attendance, participation, grades, or a combination thereof, and updating the predictive output.

21. The method of claim 12, wherein the prospect is a prospective buyer.

22. The method of claim 21, which further comprises assigning the prospective buyer to the selected agent.

23. The method of claim 22, wherein the agent history comprises history of an agent with the prospective buyer'"'"'s personality type.

24. A non-transitory computer readable medium comprising a plurality of instructions comprising:

instructions, that when executed, receive a prediction request from a user interface application on a client device;

instructions, that when executed, record a communication between an agent and a prospect on the client device and a telecommunication device of the prospect, the communication having a density of keywords;

instructions, that when executed, transcribe the recorded communication from audio to text when the recorded communication is not in text form;

instructions, that when executed, determine which segments of the text of the recorded communication originated with the agent, and which segments of the text of the recorded communication originated with the prospect;

instructions, that when executed, search for keywords and phrases in the text of the recorded communication to determine a type of communication, the density of keywords used by a prospect relating to an organization, business, service, product, or any combination thereof, and a relationship of these keywords to each other during the recorded communication, wherein the keywords are selected from a library and are related to the type of communication;

instructions, that when executed, apply a linguistic-based psychological behavioral model to the text of the recorded communication to determine a personality type of the agent and the prospect based on identifying keywords relevant to the applied linguistic-based psychological behavioral model;

instructions, that when executed, separate agent voice data from the text of the recorded communication;

instructions, that when executed, calculate a weight of keywords identified in a first segment of the agent voice data relative to other keywords identified in the first segment or a second segment in close proximity to the first segment;

instructions, that when executed, calculate an amount of time in the agent voice data that the agent used the keywords relating to the organization, business, service, or product, or any combination thereof;

instructions, that when executed, determine whether the calculated amount of time exceeds a threshold time, wherein the threshold time is based on the type of communication;

instructions, that when executed, generate a predictive output for the user interface application on the client device based on the determined personality types of the agent and the prospect, the type of communication, and the calculated weighting of the identified keywords, when the calculated amount of time exceeded the threshold time;

instructions, that when executed, access a database storing agent history with a set of prospects having the determined personality type of the prospect;

instructions, that when executed, select an agent based on the agent history with the set of prospects having the determined personality type of the prospect; and

instructions, that when executed, connect, via the IP/PBX circuit, the prospect on a subsequent communication to a workstation of the selected agent.

25. The non-transitory computer readable medium of claim 24, wherein the prospect is a prospective student or a prospective buyer.

26. The non-transitory computer readable medium of claim 24, further comprising instructions, that when executed, determine a number of meaningful contacts made between the agent and the prospect.

27. The non-transitory computer readable medium of claim 26, wherein the number of meaningful contacts is an additional basis for the generation of the predictive output.

28. An apparatus for routing outbound communications to a prospect based on a prior communication, which comprises:

an Internet Protocol Private Branch Exchange (IP/PBX) circuit;

a storage device storing a computer readable program; and

a processor executing the computer readable program comprising;

a database module adapted to;

receive a prediction request from a user interface application on a client device; and

record a communication between an agent and a prospect on the client device and a telecommunication device of the prospect, the communication having a density of keywords;

a scoring module adapted to;

transcribe the recorded communication from audio to text when the recorded communication is not in text form;

determine which segments of the text of the recorded communication originated with the agent, and which segments of the text of the recorded communication originated with the prospect;

apply at least one linguistic algorithm to determine a density of keywords relating to an organization, business, service or product, or any combination thereof, and a relationship of these keywords to each other during the recorded communication, in the text of the recorded communication to determine if the recorded communication is a meaningful contact by determining amount of time that is spent by the agent promoting a value of an organization, business, service, product, or any combination thereof, to the prospect during the recorded communication, wherein the keywords are selected from a library and are related to a type of communication;

apply a linguistic-based psychological behavioral model to the text of the recorded communication to determine a personality type of the agent and the prospect based on identifying keywords relevant to the applied linguistic-based psychological behavioral model;

separate agent voice data from the text of the recorded communication;

calculate a weight of keywords identified in a first segment of the agent voice data relative to other keywords identified in the first segment or a second segment in close proximity to the first segment;

calculate the amount of time in the agent voice data that the agent used the keywords relating to the organization, business, service, or product, or any combination thereof; and

determine whether the calculated amount of time exceeds a threshold time;

a predictive module adapted to;

generate a predictive output for the user interface application on the client device based on an output from the scoring module including the determination that the contact was meaningful and if so, the determination of the personality types for the agent and the prospect, the type of communication, and the calculated weighting of the identified keywords, when the calculated amount of time exceeded the threshold time, wherein the threshold time is based on the type of communication;

access a database storing agent history with a set of prospects having the determined personality type of the prospect;

select an agent based on the agent history with the set of prospects having the determined personality type of the prospect; and

connect, via the IP/PBX circuit, the prospect on a subsequent communication to a workstation of the selected agent.