# US 2010/0169245 A1 · Statistical Machine Learning

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

- Number: US 2010/0169245 A1
- Document ID: US20100169245A1
- Status: Published application
- Kind: A1
- Application number: 12/347,667
- 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/US20100169245A1

## CPC

- G06N 20/00: Machine learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Statistical machine learning, in which an input module receives user input that defines a hypothesis associated with a particular Output. The hypothesis defines one or more starting criteria that are proposed as being correlated with the particular output, and a recommendation engine initially provides recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis. An experience analytics system receives feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful and modifies the hypothesis based on the feedback data. Subsequent to the experience analytics system modifying the hypothesis, the recommendation engine provides recommendations that include the particular output based on the modified hypothesis.

## Claims

### Claim 1

A statistical machine learning system comprising:
an input module configured to receive user input that defines a hypothesis associated with a particular output, the hypothesis defining one or more starting criteria that are proposed as being correlated with the particular output
at least one electronic data store configured to store data representative of the hypothesis defined by the user input;
a recommendation engine configured to access the data representative of the hypothesis and provide, during a first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis; and
an experience analytics system configured to receive, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful and modify the hypothesis based on the feedback data received from the application system,
wherein the recommendation engine is further configured to, subsequent to the experience analytics system modifying the hypothesis, provide, during a second period of time that is subsequent to the first period of time, recommendations that include the particular output based on the modified hypothesis.

### Claim 2 (depends on 1)

The system of claim 1 wherein the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by determining whether the one or more starting criteria are correlated with the particular output based on the feedback data received from the application system.

### Claim 3 (depends on 2)

The system of claim 2 wherein:
the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by, in response to a determination that a particular combination of starting criteria is correlated with the particular output, using the particular combination of criteria; and
the recommendation engine is configured to provide recommendations that include the particular output based on the modified hypothesis by providing recommendations using the particular combination of criteria.

### Claim 4 (depends on 2)

The system of claim 2 wherein:
the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by, in response to a determination that a particular starting criterion is not correlated with the particular output, removing the particular starting criterion from the hypothesis; and
the recommendation engine is configured to provide recommendations that include the particular output based on the modified hypothesis by providing recommendations without regard for the particular starting criterion removed from the hypothesis.

### Claim 5 (depends on 1)

The system of claim 1 wherein:
the input module is further configured to receive user input that defines one or more declarative rules associated with the particular output, the one or more declarative rules defining one or more required criteria that is different than the one or more starting criteria defined by the hypothesis and that is required to be met in order to provide a recommendation including the particular output, and
the recommendation engine is configured to provide recommendations by providing recommendations that include the particular output conditioned on the one or more required criteria being met.

### Claim 6 (depends on 5)

The system of claim 5 wherein the recommendation engine is configured to provide recommendations that include the particular output conditioned on the one or more required criteria being met by providing recommendations based on the one or more starting criteria defined by the hypothesis only when the one or more required criteria are met.

### Claim 7 (depends on 1)

The system of claim 1 wherein:
the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by:
identifying, from the received feedback data, feedback data related to the one or more starting criteria defined by the hypothesis;
analyzing the identified feedback data related to the one or more starting criteria defined by the hypothesis;
based on the analysis, determining whether the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis;
in response to a determination that the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, establishing a rule for providing recommendations associated with the particular output based on the analysis; and
in response to a determination that the identified feedback data is not sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, continuing to monitor feedback data that is received from the application system and that is related to whether recommendations provided based on the one or more starting criteria defined by the hypothesis were successful, and
the recommendation engine is configured to provide recommendations that include the particular output based on the modified hypothesis by, in response to a determination that the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, providing recommendations that include the particular output based on the established rule.

### Claim 8 (depends on 1)

The system of claim 1 wherein:
the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by:
accessing a single instance of feedback data;
identifying one or more criteria in the single instance of feedback data that match the one or more starting criteria defined by the hypothesis; and
incrementally modifying the hypothesis based oil the one or more criteria in the single instance of feedback data that match the one or more starting criteria defined by the hypothesis, and
the recommendation engine is configured to provide recommendations that include the particular output based on the modified hypothesis by providing a next recommendation based on the incrementally-modified hypothesis.

### Claim 9 (depends on 1)

The system of claim 1 wherein:
the input module is configured to receive user input that defines a hypothesis associated with a particular output by receiving user input that defines a hypothesis associated with a particular offer, the hypothesis defining one or more starting user attributes that are proposed as being correlated with user acceptance of the particular offer;
the recommendation engine is configured to provide, during the first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis by:
accessing attributes of a user available to receive an offer;
comparing the accessed attributes of the user to the one or more starting user attributes defined by the hypothesis;
based on comparison results, determining whether at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis; and
providing the particular offer to the user in response to a determination that at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis;
the experience analytics system is configured to receive, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful by receiving, from an application system, feedback data related to whether the user accepted the particular offer;
the experience analytics system is configured to modify the hypothesis based on the feedback data received from the application system by modifying the hypothesis based on whether the user accepted the particular offer and the attributes of the user.

### Claim 10

A computer-implemented method of statistical machine learning comprising:
receiving user input that defines a hypothesis associated with a particular output, the hypothesis defining one or more starting criteria that are proposed as being correlated with the particular output;
storing, in electronic storage, data representative of the hypothesis defined by the user input;
providing, during a first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis;
receiving, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful;
modifying, using a processor, the hypothesis based on the feedback data received from the application system; and
subsequent to modifying the hypothesis, providing, during a second period of time that is subsequent to the first period of time, recommendations that include the particular output based on the modified hypothesis.

### Claim 11 (depends on 10)

The computer-implemented method of claim 10 wherein modifying the hypothesis based on the feedback data received from the application system comprises determining whether the one or more starting criteria are correlated with the particular output based on the feedback data received from the application system.

### Claim 12 (depends on 11)

The computer-implemented method of claim 11 wherein:
modifying the hypothesis based on the feedback data received from the application system further comprises, in response to a determination that a particular combination of starting criteria is correlated with the particular output, using the particular combination of criteria; and
providing recommendations that include the particular output based on the modified hypothesis comprises providing recommendations using the particular combination of criteria.

### Claim 13 (depends on 11)

The computer-implemented method of claim 11 wherein:
modifying the hypothesis based on the feedback data received from the application system further comprises, in response to a determination that a particular starting criterion is not correlated with the particular output, removing the particular starting criterion from the hypothesis; and
providing recommendations that include the particular output based on the modified hypothesis comprises providing recommendations without regard for the particular starting criterion removed from the hypothesis.

### Claim 14 (depends on 10)

The computer-implemented method of claim 10 further comprising:
receiving user input that defines one or more declarative rules associated with the particular output, the one or more declarative rules defining one or more required criteria that is different than the one or more starting criteria defined by the hypothesis and that is required to be met in order to provide a recommendation including the particular output,
wherein providing recommendations comprises providing recommendations that include the particular output conditioned on the one or more required criteria being met.

### Claim 15 (depends on 14)

The computer-implemented method of claim 14 wherein providing recommendations that include the particular output conditioned on the one or more required criteria being met comprises providing recommendations based on the one or more starting criteria defined by the hypothesis only when the one or more required criteria are met.

### Claim 16 (depends on 10)

The computer-implemented method of claim 10 wherein:
modifying the hypothesis based on the feedback data received from the application system:
identifying, from the received feedback data, feedback data related to the one or more starting criteria defined by the hypothesis;
analyzing the identified feedback data related to the one or more starting criteria defined by the hypothesis;
based on the analysis, determining whether the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis;
in response to a determination that the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, establishing a rule for providing recommendations associated with the particular output based on the analysis; and
in response to a determination that the identified feedback data is not sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, continuing to monitor feedback data that is received from the application system and that is related to whether recommendations provided based on the one or more starting criteria defined by the hypothesis were successful, and
providing recommendations that include the particular output based on the modified hypothesis comprises, in response to a determination that the identified feedback data is sufficient to establish a rule related to the one or more starting criteria defined by the hypothesis, providing recommendations that include the particular output based on the established rule.

### Claim 17 (depends on 10)

The computer-implemented method of claim 10 wherein:
modifying the hypothesis based on the feedback data received from the application system:
accessing a single instance of feedback data;
identifying one or more criteria in the single instance of feedback data that match the one or more starting criteria defined by the hypothesis; and
incrementally modifying the hypothesis based on the one or more criteria in the single instance of feedback data that match the one or more starting criteria defined by the hypothesis, and
providing recommendations that include the particular output based on the modified hypothesis comprises providing a next recommendation based on the incrementally-modified hypothesis.

### Claim 18 (depends on 10)

The computer-implemented method of claim 10 wherein:
receiving user input that defines a hypothesis associated with a particular output comprises receiving user input that defines a hypothesis associated with a particular offer, the hypothesis defining one or more starting user attributes that are proposed as being correlated with user acceptance of the particular offer;
providing, during the first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis comprises:
accessing attributes of a user available to receive an offer;
comparing the accessed attributes of the user to the one or more starting user attributes defined by the hypothesis;
based on comparison results, determining whether at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis: and
providing the particular offer to the user in response to a determination that at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis;
receiving, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful comprises receiving, from an application system, feedback data related to whether the user accepted the particular offer;
modifying the hypothesis based on the feedback data received from the application system comprises modifying the hypothesis based on whether the user accepted the particular offer and the attributes of the user.

### Claim 19

At least one computer-readable storage medium encoded with executable instructions that, when executed by a processor, cause the processor to perform operations comprising:
receiving user input that defines a hypothesis associated with a particular output, the hypothesis defining one or more starting criteria that are proposed as being correlated with the particular output;
storing, in electronic storage, data representative of the hypothesis defined by the user input;
providing, during a first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis;
receiving, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful;
modifying, using a processor, the hypothesis based on the feedback data received from the application system; and
subsequent to modifying the hypothesis, providing, during a second period of time that is subsequent to the first period of time, recommendations that include the particular output based on the modified hypothesis.

### Claim 20 (depends on 19)

The at least one computer-readable storage medium of claim 19 wherein:
receiving user input that defines a hypothesis associated with a particular output comprises receiving user input that defines a hypothesis associated with a particular offer, the hypothesis defining one or more starting user attributes that are proposed as being correlated with user acceptance of the particular offer;
providing, during the first period of time, recommendations that include the particular output based on the one or more starting criteria defined by the hypothesis comprises:
accessing attributes of a user available to receive an offer;
comparing the accessed attributes of the user to the one or more starting user attributes defined by the hypothesis;
based on comparison results, determining whether at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis; and
providing the particular offer to the user in response to a determination that at least a portion of the accessed attributes of the user match at least a portion of the one or more starting user attributes defined by the hypothesis;
receiving, from an application system, feedback data related to whether the recommendations provided based on the one or more starting criteria defined by the hypothesis were successful comprises receiving, from an application system, feedback data related to whether the user accepted the particular offer;
modifying the hypothesis based on the feedback data received from the application system comprises modifying the hypothesis based on whether the user accepted the particular offer and the attributes of the user.
