Patent Issued for Heuristic Context Prediction Engine (USPTO 10,970,641)

2021 APR 20 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- From Alexandria, Virginia, NewsRx journalists report that a patent by the inventors Flowers, Elizabeth (Bloomington, IL); Dua, Puneit (Bloomington, IL); Balota, Eric (Bloomington, IL); Phillips, Shanna L. (Bloomington, IL), filed on April 24, 2017, was published online on April 19, 2021.

The patent’s assignee for patent number 10,970,641 is State Farm Mutual Automobile Insurance Company (Bloomington, Illinois, United States).

News editors obtained the following quote from the background information supplied by the inventors: “Organizations involved in customer service activities often process large amounts of unstructured data to make decisions while interacting with a customer in real-time. For example, in the case of a customer service representative speaking on the telephone with a customer experiencing an issue with a product or service, appropriate solutions may include a combination of timeliness of response and accuracy in content.

“Such unstructured data may include voluminous transaction records spanning decades, unstructured customer service data, or real-time transcripts of customer service interactions with scattered contextual indicators. To reasonably expect a customer service representative to effectively leverage such large data sets in real-time places an unreasonable burden on a customer service representative. However, failing to do so robs the customer service representative of vital context not readily apparent, and the wealth of knowledge gained throughout the history of an organization that would otherwise need to be distilled to briefing materials and expensively trained over time. Thus, organizations may value tools to rapidly process large data sets, to infer context, suggest lessons learned based upon transaction data, while learning through successive process iterations. Furthermore, appropriate application of such tools may provide a competitive advantage in a crowded and competitive customer service industry.

“In an effort to automate and provide better predictability of customer service experiences, many organizations develop customer relationship management (CRM) software packages. Organizations that develop these software packages often develop custom solutions, at great expense, to best meet the needs of their customers in unique industries. Such tools while providing a great level of detail for the customer service representative, lack the flexibility to react to changing business conditions or fully exploit the underlying technology, driving additional cost into an already expensive solution.

“Some organizations where able to make concessions on customized solutions turn to off-the-shelf or commercially available software solutions that reduce the overall cost of implementation. Such solutions may provide customer service representative prompting tools with question and answer formats that allow for consistency of customer experience, however, at the expense of a less personalized experience required in many industries. While more flexible than fully-custom solutions, the impersonal question-answer format of customer interaction may not improve without costly software revisions, rarely performed by original equipment manufacturers (OEMs) of off-the-shelf solutions.

“The ability for a customer service experience to learn and improve over successive iterations remains paramount for organizations to offer discriminating customer service experiences. Often the burden of continual improvement falls to the customer service representative, as a human being able to adapt and learn to changing conditions more rapidly even within the confines of a rigid customer service software application. However, with the advent of outsourcing prevalent in the customer service industry, the customer service representative may lack much of the necessary context required to provide high levels of relevant customer service. This lack of context in an interconnected company is less an issue of distance and more an issue of data access and the ability to contextually process data to present relevant solutions in a timely manner.”

As a supplement to the background information on this patent, NewsRx correspondents also obtained the inventors’ summary information for this patent: “One exemplary embodiment includes a computer-implemented method, executed with a computer processor, to predict a current and subsequent context. The method may include retrieving an un-structured website history transaction data set stored in a first memory, receiving a unique customer identifier, accessing a heuristic algorithm, and/or executing the algorithm using the data set and the identifier. The algorithm may output a correlation score associated with at least one user and predict the current context using at least one correlation score, calculate a predicted question using the current context, and/or update the algorithm using the subsequent context. The method may include additional, less, or alternate actions, including those discussed elsewhere herein.

“Yet another exemplary embodiment includes a computer-implemented method, executed with a computer processor, that generates a predicted subsequent context using a chat window that includes retrieving an un-structured transaction set correlating questions and answers stored in a first memory, receiving a natural language input from a chat window from a customer, and/or accessing and executing a heuristic algorithm to generate the predicted subsequent context using the language input and the data set. The embodiment includes calculating a predicted question using the predicted subsequent context, receiving an actual customer question with a human machine interface, and/or updating the algorithm using a calculated correlation between the actual customer question and the predicted question. The method may include additional, less, or alternate actions, including those discussed elsewhere herein.

“An alternative embodiment includes a computer-implemented method, executed with a computer processor, that generates a predicted subsequent customer question and suggested answers. The method may include retrieving an un-structured transaction set correlating past customer questions and/or receiving a natural language input in a customer service environment. Furthermore, the method may include accessing and executing a heuristic algorithm to generate a predicted subsequent customer question using the language input and the transaction set. Still further, the embodiment may include receiving, with the processor, an actual customer question with a human machine interface and/or updating the algorithm using a calculated correlation between the actual customer question and the predicted subsequent question. The method may include additional, less, or alternate actions, including those discussed elsewhere herein.

“Another exemplary embodiment includes a computer-implemented method, executed with a computer processor, that generates an agent training suggestion using a natural language input and an unstructured agent transaction record. The method may include retrieving an un-structured agent transaction record and receiving a natural language input. The method may include accessing and executing a heuristic algorithm to generate the agent training suggestion using the transaction record and the natural language input. Furthermore, the method may include receiving an indication of agent behavior modification and/or updating the algorithm using a calculated correlation between the suggestion and the indication. The method may include additional, less, or alternate actions, including those discussed elsewhere herein.

“Exemplary embodiments may include computer-implemented methods that may in other embodiments include apparatus configured to implement the method, and/or non-transitory computer readable mediums comprising computer-executable instructions that cause a processor to perform the method.

“Advantages will become more apparent to those skilled in the art from the following description of the preferred embodiments which have been shown and described by way of illustration. As will be realized, the present embodiments may be capable of other and different embodiments, and their details are capable of modification in various respects. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive.”

The claims supplied by the inventors are:

“What is claimed is:

“1. A computer-implemented method, executed with processor, comprising: retrieving, by the processor, an un-structured website history transaction data set stored in a first memory; receiving, by the processor and from a network interface device, an identifier uniquely identifying a user of a plurality of users; accessing, by the processor, a heuristic algorithm stored in a second memory; executing the heuristic algorithm, by the processor, and using the un-structured website history transaction data set and the identifier, wherein executing the heuristic algorithm causes the heuristic algorithm to output a correlation score associated with the user; predicting, by the processor and using the correlation score, a first context indicating a user category; generating, by the processor and using the first context, a predicted question of the user; and updating, by the processor and based at least in part on the predicted question, the heuristic algorithm in the second memory using a second context received from the network interface device.

“2. The computer-implemented method of claim 1, wherein the identifier comprises an internet network address.

“3. The computer-implemented method of claim 1, wherein the identifier comprises a source telephone number.

“4. The computer-implemented method of claim 1, wherein the un-structured website history transaction data set comprises past transactions related to at least one account.

“5. The computer-implemented method of claim 1, wherein the first memory comprises an external transaction server.

“6. The computer-implemented method of claim 1, wherein the second memory comprises an external heuristic server.

“7. The computer-implemented method of claim 1, wherein the un-structured website history transaction data set corresponds to the identifier provided via a human-machine interface.

“8. The computer-implemented method of claim 1, further comprising: receiving, by the processor, the second context, wherein the second context comprises a question from the user; determining, by the processor, a difference between the predicted question and the second context; and updating, by the processor and in response to determining the difference, the heuristic algorithm in the second memory.

“9. The computer-implemented method of claim 1, further comprising: receiving a question from the user; and predicting, based at least in part on the question, the first context.

“10. A computer system comprising one or more processors configured to: retrieve an un-structured website history transaction data set stored in a first memory; receive, from a network interface device, an identifier uniquely identifying a user of a plurality of users; access a heuristic algorithm stored in a second memory; execute the heuristic algorithm, using the un-structured website history transaction data set and the identifier, wherein executing the heuristic algorithm causes the heuristic algorithm to output a correlation score associated with the user; predict, using the correlation score, a first context indicating a user category; generate, using the first context, a predicted question of the user; and update, based at least in part on the predicted question, the heuristic algorithm in the second memory using a second context received from the network interface device.

“11. The computer system of claim 10, wherein the identifier comprises an internet network address.

“12. The computer system of claim 10, wherein the identifier comprises a source telephone number.

“13. The computer system of claim 10, wherein the un-structured website history transaction data set comprises past transactions related to at least one account.

“14. The computer system of claim 10, wherein the first memory comprises an external transaction server, and the second memory comprises an external heuristic server.

“15. A non-transitory computer readable medium, comprising computer readable instructions that, when executed by processor, cause the processor to perform acts comprising: retrieving an un-structured website history transaction data set stored in a first memory; receiving, from a network interface device, an identifier uniquely identifying a user of a plurality of users; accessing a heuristic algorithm stored in a second memory; executing the heuristic algorithm, using the un-structured website history transaction data set and the identifier, wherein executing the heuristic algorithm causes the heuristic algorithm to output a correlation score associated with the user; predicting, using the correlation score, a first context indicating a user category; generating, using the first context, a predicted question of the user; and updating, based at least in part on the predicted question, the heuristic algorithm in the second memory using a second context received from the network interface device.

“16. The non-transitory computer readable medium of claim 15, wherein the identifier comprises an internet network address.

“17. The non-transitory computer readable medium of claim 15, wherein the identifier comprises a source telephone number.

“18. The non-transitory computer readable medium of claim 15, wherein the un-structured website history transaction data set comprises past transactions related to at least one account.

“19. The non-transitory computer readable medium of claim 15, wherein the first memory comprises an external transaction server, and the second memory comprises an external heuristic server.

“20. The non-transitory computer readable medium of claim 15, wherein the un-structured website history transaction data set corresponds to the identifier provided via a human-machine interface.”

For additional information on this patent, see: Flowers, Elizabeth; Dua, Puneit; Balota, Eric; Phillips, Shanna L. Heuristic Context Prediction Engine. U.S. Patent Number 10,970,641, filed April 24, 2017, and published online on April 19, 2021. Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=10,970,641.PN.&OS=PN/10,970,641RS=PN/10,970,641

(Our reports deliver fact-based news of research and discoveries from around the world.)