Granted Patent B2

US 12,405,972 · App. 17/973,311 · Granted Sep 2, 2025

Inventors: Jan Portisch (Bruchsal, DE); Sandra Bracholdt (Dielheim, DE)

Assignee: SAP SE

G06F16/288G06F16/2228G06F16/2477

Abstract

Claims (78)

1. A computing system for utilizing relationship data, the computing system comprising:

at least one processor programmed to perform operations comprising:

receiving, from a second computing system, a first time-dependent query against a knowledge graph data structure, the first time-dependent query comprising an indication of a first test relationship, and an indication of a first test point-in-time, the first test relationship being described by a first time period during which the first test relationship is true;

accessing confirmation data from the knowledge graph data structure, the confirmation data describing a first plurality of confirmation points-in-time at which the first test relationship is true;

determining that at least one of a beginning or an end of the first time period is not defined by the knowledge graph data structure at least in part by determining that the first plurality of confirmation points-in-time do not extend past at least one of the beginning or the end of the first time period;

determining a response to the first time-dependent query indicating a veracity of the first test relationship at the first test point-in-time using the first plurality of confirmation points-in-time; and

sending the response to the first time-dependent query to the second computing system.

2. The computing system of claim 1 , the operations further comprising:

determining that the beginning of the first time period is defined and that the end of the first time period is not defined by the knowledge graph data structure; and

determining that a number of the first plurality of confirmation points-in-time are before the first test point-in-time, the response to the first time-dependent query indicating that the first test relationship is true at the first test point-in-time.

3. The computing system of claim 1 , the operations further comprising:

determining that the beginning of the first time period is defined and that the end of the first time period is not defined by the knowledge graph data structure; and

fitting a probability distribution to the first plurality of confirmation points-in-time, the veracity of the first test relationship at the first test point-in-time being a value of the probability distribution at the first test point-in-time.

4. The computing system of claim 1 , the operations further comprising:

determining that the beginning of the first time period is not defined by the knowledge graph data structure;

determining that the end of the first time period is not defined by the knowledge graph data structure; and

positioning a probability distribution using the first plurality of confirmation points-in-time, the veracity of the first test relationship at the first test point-in-time being a value of the probability distribution at the first test point-in-time.

5. The computing system of claim 1 , the operations further comprising:

determining that the first test relationship is a multi-period relationship;

selecting a first confirmation point-in-time of the first plurality of confirmation points-in-time that is after the first test point-in-time; and

selecting a second confirmation point-in-time of the first plurality of confirmation points-in-time that is before the first test point-in-time, the veracity of the first test relationship at the first test point-in-time being based at least in part on a first probability distribution centered at the first confirmation point-in-time and a second probability distribution centered on the second confirmation point-in-time.

6. The computing system of claim 1 , the accessing of the confirmation data comprising accessing a first triple data unit from the knowledge graph data structure, the first triple data unit corresponding to the first test relationship.

7. The computing system of claim 1 , the accessing of the confirmation data comprising accessing a plurality of triple data units from the knowledge graph data structure.

8. The computing system of claim 1 , the operations further comprising:

accessing a candidate relationship comprising a candidate subject entity, a candidate predicate, a candidate object, and a candidate point-in-time;

accessing a second plurality of confirmation points-in-time at which the candidate relationship is true;

determining veracity of the candidate relationship at the candidate point-in-time using the second plurality of confirmation points-in-time;

determining that the veracity of the candidate relationship at the candidate point-in-time is greater than a threshold value; and

after determining that the veracity of the candidate relationship at the candidate point-in- time is greater than the threshold value, incorporating the candidate relationship into the knowledge graph data structure.

9. The computing system of claim 1 , the operations further comprising:

accessing a second test triple data unit from the knowledge graph data structure, the second test triple data unit describing a second test relationship and a second test point-in-time;

accessing second confirmation data from the knowledge graph data structure, the second confirmation data describing a second plurality of confirmation points-in-time at which the second test relationship is true;

determining that a veracity of the second test relationship at the second test point-in-time is less than a threshold value; and

obtaining a confirmation of the second test relationship.

10. The computing system of claim 1 , the knowledge graph data structure being stored at a first tenancy of a plurality of tenancies at a cloud deployment, the operations further comprising:

accessing a second knowledge graph data structure stored at a second tenancy of the plurality of tenancies at the cloud deployment; and

using the second knowledge graph data structure to determine at least a portion of the first plurality of confirmation points-in-time at which the first test relationship is true.

11. A method for utilizing relationship data in a knowledge graph computing system, the method comprising:

receiving, by the knowledge graph computing system and from a second computing system, a first time-dependent query against a knowledge graph data structure, the first time-dependent query comprising an indication of a first test relationship and an indication of a first test point-in-time, the first test relationship being described by a first time period during which the first test relationship is true;

accessing, by the knowledge graph computing system, confirmation data from the knowledge graph data structure, the confirmation data describing a first plurality of confirmation points-in-time at which the first test relationship is true;

determining that at least one of a beginning or an end of the first time period is not defined by the knowledge graph data structure at least in part by determining that the first plurality of confirmation points-in-time do not extend past at least one of the beginning or the end of the first time period;

determining, by the knowledge graph computing system, a response to the first time-dependent query indicating a veracity of the first test relationship at the first test point-in-time using the first plurality of confirmation points-in-time; and

sending the response to the first time-dependent query, by the knowledge graph computing system and to the second computing system.

12. The method of claim 11 , further comprising:

determining, by the knowledge graph computing system, that the beginning of the first time period is defined and that the end of the first time period is not defined by the knowledge graph data structure; and

determining, by the knowledge graph computing system, that a number of the first plurality of confirmation points-in-time are before the first test point-in-time, the response to the first time-dependent query indicating that the first test relationship is true at the first test point-in-time.

13. The method of claim 11 , further comprising:

determining, by the knowledge graph computing system, that the beginning of the first time period is defined and that the end of the first time period is not defined by the knowledge graph data structure; and

fitting a probability distribution to the first plurality of confirmation points-in-time, the veracity of the first test relationship at the first test point-in-time being a value of the probability distribution at the first test point-in-time.

14. The method of claim 11 , further comprising:

determining, by the knowledge graph computing system, that the beginning of the first time period is not defined by the knowledge graph data structure;

determining, by the knowledge graph computing system, that the end of the first time period is not defined by the knowledge graph data structure; and

positioning a probability distribution using the first plurality of confirmation points-in-time, the veracity of the first test relationship at the first test point-in-time being a value of the probability distribution at the first test point-in-time.

15. The method of claim 11 , further comprising:

determining, by the knowledge graph computing system, that the first test relationship is a multi-period relationship;

selecting, by the knowledge graph computing system, a first confirmation point-in-time of the first plurality of confirmation points-in-time that is after the first test point-in-time; and

selecting, by the knowledge graph computing system, a second confirmation point-in-time of the first plurality of confirmation points-in-time that is before the first test point-in-time, the veracity of the first test relationship at the first test point-in-time being based at least in part on a first probability distribution centered at the first confirmation point-in-time and a second probability distribution centered on the second confirmation point-in-time.

16. The method of claim 11 , the accessing of the confirmation data comprising accessing a first triple data unit from the knowledge graph data structure, the first triple data unit corresponding to the first test relationship.

17. The method of claim 11 , further comprising:

accessing, by the knowledge graph computing system, a candidate relationship comprising a candidate subject entity, a candidate predicate, a candidate object, and a candidate point-in-time;

accessing, by the knowledge graph computing system, a second plurality of confirmation points-in-time at which the candidate relationship is true;

determining, by the knowledge graph computing system, a veracity of the candidate relationship at the candidate point-in-time using the second plurality of confirmation points-in-time;

determining, by the knowledge graph computing system, that the veracity of the candidate relationship at the candidate point-in-time is greater than a threshold value; and

after determining that the veracity of the candidate relationship at the candidate point-in-time is greater than the threshold value, incorporating the candidate relationship into the knowledge graph data structure.

18. The method of claim 11 , further comprising:

accessing, by the knowledge graph computing system, a second test triple data unit from the knowledge graph data structure, the second test triple data unit describing a second test relationship and a second test point-in-time;

accessing, by the knowledge graph computing system, second confirmation data from the knowledge graph data structure, the second confirmation data describing a second plurality of confirmation points-in-time at which the second test relationship is true;

determining, by the knowledge graph computing system, that a veracity of the second test relationship at the second test point-in-time is less than a threshold value; and

obtaining, by the knowledge graph computing system, a confirmation of the second test relationship.

19. The method of claim 11 , the knowledge graph data structure being stored at a first tenancy of a plurality of tenancies at a cloud deployment, the method further comprising:

accessing, by the knowledge graph computing system, a second knowledge graph data structure stored at a second tenancy of the plurality of tenancies at the cloud deployment; and

using, by the knowledge graph computing system, the second knowledge graph data structure to determine at least a portion of the first plurality of confirmation points-in-time at which the first test relationship is true.

20. A machine-readable medium comprising instructions thereon that, when executed by at least one processor, cause the at least one processor to perform operations comprising:

receiving, from a second computing system, a first time-dependent query against a knowledge graph data structure, the first time-dependent query comprising an indication of a first test relationship, and an indication of a first test point-in-time, the first test relationship being described by a first time period during which the first test relationship is true;

accessing confirmation data from the knowledge graph data structure, the confirmation data describing a first plurality of confirmation points-in-time at which the first test relationship is true;

determining that at least one of a beginning or an end of the first time period is not defined by the knowledge graph data structure at least in part by determining that the first plurality of confirmation points-in-time do not extend past at least one of the beginning or the end of the first time period;

determining a response to the first time-dependent query indicating a veracity of the first test relationship at the first test point-in-time using the first plurality of confirmation points-in-time; and

sending the response to the first time-dependent query to the second computing system.

Assignments (1)

Continuity (2)

References Cited (10)

Cited By (1)

US 12,541,693