# US 2010/0169067 A1 · Optimization Technology

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0169067 A1
- Document ID: US20100169067A1
- Status: Published application
- Kind: A1
- Application number: 12/347,649
- Filed: 2008-12-31
- Published: 2010-07-01
- Earliest claimed: 2008-12-31
- Assignees: Sap SE
- Inventors: Binyamin Basel; Yizhak Idan
- Web page: https://patentgrep.com/patent/US20100169067A1

## CPC

- G06F 13/10: Program control for peripheral devices (G06F13/14 - G06F13/42 take precedence)
- G06Q 30/02: Marketing; Price estimation or determination; Fundraising
- G06Q 30/0255: {based on user history}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR

## Abstract

Predictive modeling, in which a first simulation model that simulates a first policy for managing a set of data is determined based on a pattern representative of the set of data and first operating variables for the first policy. The first simulation model is optimized and first prediction data that estimates performance of the first policy is computed using the optimized first simulation model. A second simulation model that simulates a second policy for managing the set of data is determined based on the pattern and second operating variables for the second policy. The second simulation model is optimized and second prediction data that estimates performance of the second policy is computed using the optimized second simulation model. The first prediction data is compared with the second prediction data, and a determination of whether the first policy is better than the second policy is made.

## Claims

### Claim 1

A system comprising:
an optimization and recommendation engine configured to identify an interaction with a user that presents an opportunity to provide a recommendation to the user;
an electronic data store configured to store recommendation data associated with available recommendations;
a qualification system configured to access, from the electronic data store, recommendation data associated with available recommendations and perform qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified; and
a cache configured to cache user data associated with the user corresponding to the identified interaction and cache the identified recommendations for which the user is qualified,
wherein the optimization and recommendation engine is further configured to:
determine initial interaction information related to the identified interaction with the user;
select, from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information;
provide the first recommendation for presentation to the user;
monitor interactions relevant to the identified interaction with the user;
based on the monitoring, detect a change in interaction information related to the identified interaction with the user;
in response to detecting the change in interaction information, select, from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information, the second recommendation being different than the first recommendation; and
provide the second recommendation for presentation to the user.

### Claim 2

A computer-implemented method of optimizing recommendations, the method comprising using one or more processors to perform operations comprising:
identifying an interaction with a user that presents an opportunity to provide a recommendation to the user;
loading, into a cache, user data associated with the user corresponding to the identified interaction;
accessing, from an electronic data store, recommendation data associated with available recommendations;
performing qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified;
loading, into the cache, the identified recommendations for which the user is qualified;
determining initial interaction information related to the identified interaction with the user;
selecting, from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information;
providing the first recommendation for presentation to the user;
monitoring interactions relevant to the identified interaction with the user;
based on the monitoring, detecting a change in interaction information related to the identified interaction with the user;
in response to detecting the change in interaction information, selecting, from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information, the second recommendation being different than the first recommendation; and
providing the second recommendation for presentation to the user.

### Claim 3 (depends on 2)

The computer-implemented method of claim 2 wherein loading, into a cache, user data associated with the user corresponding to the identified interaction comprises, in response to identifying the interaction with the user:
accessing, from the electronic data store, user data associated with the user corresponding to the identified interaction;
indexing the accessed user data; and
loading, into the cache, the indexed user data.

### Claim 4 (depends on 2)

The computer-implemented method of claim 2 wherein loading, into a cache, user data associated with the user corresponding to the identified interaction comprises loading, into the cache, user data associated with the user corresponding to the identified interaction prior to identifying the interaction with the user.

### Claim 5 (depends on 2)

The computer-implemented method of claim 2 wherein performing qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified comprises identifying, based on the user data loaded into the cache, recommendations for which the user meets eligibility criteria.

### Claim 6 (depends on 2)

The computer-implemented method of claim 2 wherein performing qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified comprises:
identifying an agent handling the interaction with the user;
accessing, from the electronic data store, skills information associated with the identified agent;
identifying, based on the accessed skills information, recommendations that correspond to the skills of the identified agent handling the interaction with the user.

### Claim 7 (depends on 2)

The computer-implemented method of claim 2 wherein performing qualification analysis on the accessed recommendation data to identity recommendations for which the user corresponding to the identified interaction is qualified comprises:
determining, based on the user data loaded into the cache, a target profile corresponding to the user: and
identifying recommendations that correspond to the target profile corresponding to the user.

### Claim 8 (depends on 2)

The computer-implemented method of claim 2 wherein performing qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified comprises:
identifying, based on the user data loaded into the cache, recommendations previously offered to the user; and
removing recommendations that have been previously offered to the user.

### Claim 9 (depends on 2)

The computer-implemented method of claim 2 wherein performing qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified comprises filtering the accessed recommendation data through a series of qualification gates including (1) a first qualification gate that filters recommendations for which the user does not meet eligibility criteria, (2) a second qualification gate that filters recommendations for which an agent handling the interaction with the user does not possess skills necessary to effectively present the recommendations, (3) a third qualification gate that filters recommendations that do not correspond to a target profile associated with the user, and (4) a fourth qualification gate that filters recommendations that have been previously offered to the user.

### Claim 10 (depends on 2)

The computer-implemented method of claim 2 selecting, from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information comprises selecting the first recommendation based on a monetary value corresponding to the first recommendation.

### Claim 11 (depends on 2)

The computer-implemented method of claim 2 selecting, from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information comprises:
determining, for each of the identified recommendations loaded into the cache, a user suitability score based on the user data loaded into the cache and the determined initial interaction information;
determining, for each of the identified recommendations loaded into the cache, a value score that reflects a relative monetary value of the corresponding recommendation;
determining, for each of the identified recommendations loaded into the cache, a combined score based on the determined user suitability scores and the determined value scores, the combined scores reflecting a combination of expected response to the recommendations and business value of the recommendations; and
selecting the first recommendation based on the determined combined scores that reflect a combination of expected response to the recommendations and business value of the recommendations.

### Claim 12 (depends on 11)

The computer-implemented method of claim 11 wherein selecting the first recommendation based on the determined combined scores comprises:
determining that at least two recommendations have the combined score that reflects the highest combination of expected response and business value of the identified recommendations loaded into the cache;
comparing the user suitability scores associated with the at least two recommendations; and
selecting, from among the at least two recommendations, the recommendation having the user suitability score that reflects the highest expectation of response.

### Claim 13 (depends on 2)

The computer-implemented method of claim 2 wherein:
providing the first recommendation for presentation to the user comprises sending, over a network, the first recommendation to an application system; and
monitoring interactions relevant to the identified interaction with the user comprises receiving, over the network and from the application system, feedback related to the user's reaction to the first recommendation.

### Claim 14 (depends on 2)

The computer-implemented method of claim 2 wherein:
detecting a change in interaction information related to the identified interaction with the user comprises detecting that the user moves to a different section of a website while browsing the website; and
selecting, from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information comprises selecting, from among the identified recommendations loaded into the cache, a second recommendation that is related to the different section of the website.

### Claim 15 (depends on 2)

The computer-implemented method of claim 2 wherein:
detecting a change in interaction information related to the identified interaction with the user comprises detecting that the user has made a complaint about price of a good or service during the interaction;
selecting, from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information comprises excluding recommendations associated with a good or service that has a relatively high price.

### Claim 16 (depends on 2)

The computer-implemented method of claim 2 wherein:
monitoring interactions relevant to the identified interaction with the user comprises monitoring aspects of service being provided in an environment in which the identified interaction with the user is occurring,
detecting a change in interaction information related to the identified interaction with the user comprises detecting a change in the service being provided in an environment in which the identified interaction with the user is occurring;
selecting, from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information comprises selecting a second recommendation based on the detected change in the service being provided in an environment in which the identified interaction with the user is occurring.

### Claim 17 (depends on 2)

The computer-implemented method of claim 2 wherein:
determining initial interaction information related to the identified interaction with the user comprises determining a communication channel associated with the interaction with the user and determining a reason that the user initiated the interaction; and
selecting, from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information comprises selecting a first recommendation that is suitable for the communication channel associated with the interaction with the user and that accounts for the reason that the user initiated the interaction.

### Claim 18 (depends on 2)

The computer-implemented method of claim 2 further comprising:
in response to determining the initial interaction information related to the identified interaction with the user, loading, into the cache, the initial interaction information;
in response to providing the first recommendation for presentation to the user, updating, in the cache, the interaction information to indicate that the first recommendation has been presented to the user;
in response to detecting the change in interaction information, updating, in the cache, the interaction information to reflect the detected change in interaction information; and
in response to providing the second recommendation for presentation to the user, updating, in the cache, the interaction information to indicate that the second recommendation has been presented to the user.

### Claim 19 (depends on 2)

The computer-implemented method of claim 2 further comprising:
identifying an agent handling the interaction with the user:
loading, into the cache, agent data associated with the agent handling the interaction with the user; and
selecting, from among the identified recommendations loaded into the cache, a first recommendation further comprises selecting, from among the identified recommendations loaded into the cache, a first recommendation based on the agent data loaded into the cache.

### Claim 20

At least one computer-readable storage medium encoded with executable instructions that, when executed by a processor, cause the processor to perform operations comprising:
identifying, using a processor, an interaction with a user that presents an opportunity to provide a recommendation to the user;
loading, into a cache, user data associated with the user corresponding to the identified interaction;
accessing, from an electronic data store, recommendation data associated with available recommendations;
performing, using the processor, qualification analysis on the accessed recommendation data to identify recommendations for which the user corresponding to the identified interaction is qualified;
loading, into the cache, the identified recommendations for which the user is qualified;
determining, using the processor, initial interaction information related to the identified interaction with the user;
selecting, using the processor and from among the identified recommendations loaded into the cache, a first recommendation based on the user data loaded into the cache and the determined initial interaction information;
providing, using the processor, the first recommendation for presentation to the user;
monitoring, using the processor, interactions relevant to the identified interaction with the user;
based on the monitoring, detecting, using the processor, a change in interaction information related to the identified interaction with the user;
in response to detecting the change in interaction information, selecting, using the processor and from among the identified recommendations loaded into the cache, a second recommendation based on the user data loaded into the cache and the change in interaction information, the second recommendation being different than the first recommendation; and
providing, using the processor, the second recommendation for presentation to the user.
