# US 2024/0202212 A1 · Contextualizing data to augment processes using semantic technologies and artificial intelligence

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

- Number: US 2024/0202212 A1
- Document ID: US20240202212A1
- Status: Published application
- Kind: A1
- Application number: 18/083,024
- Filed: 2022-12-16
- Published: 2024-06-20
- Earliest claimed: 2022-12-16
- Assignees: Sap SE
- Applicants: Sap SE
- Inventors: Gilad Vidal; Marcus Krug
- Web page: https://patentgrep.com/patent/US20240202212A1

## CPC

- G06F 16/283: {Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP}
- G06F 16/2455: Query execution
- G06N 5/022: Knowledge engineering; Knowledge acquisition
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

The present disclosure involves systems, software, and computer implemented methods for contextualizing data to augment processes using semantic technologies and artificial intelligence. One example method includes identifying one or more data sources for semantic analysis. The data sources can include a data warehouse, a database, or a data lake. User behaviors of one or more users are identified for semantic analysis. The user behaviors include behaviors of how the users consume data in the one or more data sources. A semantic model is generated, using a knowledge graph, for the user behaviors. Nodes of the knowledge graph correspond to a class of entities in the data sources and are annotated with user behaviors and data source information. One or more queries from a user are monitored. Data from at least one node of the knowledge graph is recommended to the user, based on the semantic model and the queries.

## Claims

### Claim 1

A computer-implemented method, comprising:
identifying one or more data sources for semantic analysis, wherein the one or more data sources comprise at least one of a data warehouse, a database, or a data lake;
identifying one or more user behaviors of one or more users for semantic analysis, wherein the one or more user behaviors comprise behaviors of how the one or more users consume data in the one or more data sources;
generating, using a knowledge graph, a semantic model for the one or more user behaviors, wherein nodes of the knowledge graph correspond to a class of entities in the one or more data sources and are annotated with user behaviors and data source information;
monitoring one or more queries from a user; and
recommending, based on the semantic model and the one or more queries from the user, data from at least one node of the knowledge graph to the user.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein user behaviors comprise a user report request, a user query, or logged information indicating user activity with one or more processes.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, wherein a user behavior using a data source comprises a user context.

### Claim 4 (depends on 3)

The computer-implemented method of claim 3, further comprising computing relevance scores for data sources for user contexts, wherein a respective relevance score for a data source for a respective user context indicate a strength of match of the data source to the user context.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4, wherein recommending data from at least one knowledge graph node to the user comprises identifying one or more nodes that are associated with data sources have highest relevance scores to a user context that matches the one or more queries of the user.

### Claim 6 (depends on 4)

The computer-implemented method of claim 4, further comprising adding data sources with highest relevance scores to a data catalog.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, further comprising:
receiving a search request from a user for data in the data catalog that matches a user context;
identifying catalog entries in the data catalog that match the user context; and
providing, in response to the search request, the identified catalog entries that match the user context of the search request.

### Claim 8 (depends on 1)

The computer-implemented method of claim 1, wherein the one or more data sources include structured and unstructured data.

### Claim 9 (depends on 1)

The computer-implemented method of claim 1, wherein the semantic model is developed using one or more of natural language processing and machine learning.

### Claim 10 (depends on 1)

The computer-implemented method of claim 1, further comprising:
determining whether the user interacted with recommended data; and
updating the semantic model based on whether the user interacted with recommended data.

### Claim 11 (depends on 1)

The computer-implemented method of claim 1, further comprising:
identifying a new data source;
performing semantic analysis on the new data source; and
updating the semantic model based on the semantic analysis of the new data source including adding or updating nodes in the knowledge graph to include data from the new data source.

### Claim 12 (depends on 11)

The computer-implemented method of claim 11, further comprising determining relevance scores for added and updated nodes for user contexts in the semantic model.

### Claim 13 (depends on 12)

The computer-implemented method of claim 12, further comprising recommending knowledge graph data that includes data from the new data source as a match to a user context of a new user query.

### Claim 14 (depends on 1)

The computer-implemented method of claim 1, wherein recommending data from at least one node of the knowledge graph to the user comprises recommending, to the user, a data source associated with a first node.

### Claim 15

A system comprising:
one or more computers; and
a computer-readable medium coupled to the one or more computers having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising:
identifying one or more data sources for semantic analysis, wherein the one or more data sources comprise at least one of a data warehouse, a database, or a data lake;
identifying one or more user behaviors of one or more users for semantic analysis, wherein the one or more user behaviors comprise behaviors of how the one or more users consume data in the one or more data sources;
generating, using a knowledge graph, a semantic model for the one or more user behaviors, wherein nodes of the knowledge graph correspond to a class of entities in the one or more data sources and are annotated with user behaviors and data source information;
monitoring one or more queries from a user; and
recommending, based on the semantic model and the one or more queries from the user, data from at least one node of the knowledge graph to the user.

### Claim 16 (depends on 15)

The system of claim 15, wherein user behaviors comprise a user report request, a user query, or logged information indicating user activity with one or more processes.

### Claim 17 (depends on 15)

The system of claim 15, wherein the operations further comprise computing relevance scores for data sources for user contexts, wherein a respective relevance score for a data source for a respective user context indicate a strength of match of the data source to the user context.

### Claim 18

A non-transitory, computer-readable medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations, the operations comprising:
identifying one or more data sources for semantic analysis, wherein the one or more data sources comprise at least one of a data warehouse, a database, or a data lake;
identifying one or more user behaviors of one or more users for semantic analysis, wherein the one or more user behaviors comprise behaviors of how the one or more users consume data in the one or more data sources;
generating, using a knowledge graph, a semantic model for the one or more user behaviors, wherein nodes of the knowledge graph correspond to a class of entities in the one or more data sources and are annotated with user behaviors and data source information;
monitoring one or more queries from a user; and
recommending, based on the semantic model and the one or more queries from the user, data from at least one node of the knowledge graph to the user.

### Claim 19 (depends on 18)

The computer-readable medium of claim 18, wherein user behaviors comprise a user report request, a user query, or logged information indicating user activity with one or more processes.

### Claim 20 (depends on 18)

The computer-readable medium of claim 18, wherein the operations further comprise computing relevance scores for data sources for user contexts, wherein a respective relevance score for a data source for a respective user context indicate a strength of match of the data source to the user context.
