# US 2015/0149191 A1 · Apparatus and method for matching large-scale biomedical ontologies

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

- Number: US 2015/0149191 A1
- Document ID: US20150149191A1
- Status: Published application
- Kind: A1
- Application number: 14/091,689
- Filed: 2013-11-27
- Published: 2015-05-28
- Earliest claimed: 2013-11-27
- Assignees: University-Industry Cooperation Group of Kyung Hee University
- Inventors: Sung-Young LEE; Muhammad Bilal Amin
- Web page: https://patentgrep.com/patent/US20150149191A1

## CPC

- G06Q 10/063: Operations research, analysis or management
- G06F 40/30: Semantic analysis
- G16B 50/10: Ontologies; Annotations
- G16H 15/00: ICT specially adapted for medical reports, e.g. generation or transmission thereof
- G16H 70/00: ICT specially adapted for the handling or processing of medical references
- 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
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

An ontology matching apparatus for large-scale biomedical ontologies according to the present invention is provided, and the ontology matching apparatus includes a preprocessing unit configured to classify received candidate ontologies into one or more ontology subsets to generate ontology subsets, a distribution processing unit configured to divide the generated ontology subsets by virtue of a distribution algorithm, apply a matching algorithm to the divided ontology subsets to generate matching threads, and deliver the generated matching threads to individual cores of participating nodes, and an aggregating unit configured to collect and sum matching results generated by the individual cores performing matching operations based on the matching threads to generate an ontology mapping.

## Claims

### Claim 1

An ontology matching apparatus for large-scale biomedical ontologies, the apparatus comprising:
a preprocessing unit configured to classify received candidate ontologies into one or more ontology subsets to generate ontology subsets;
a distribution processing unit configured to divide the generated ontology subsets by a distribution algorithm, apply a matching algorithm to the divided ontology subsets to generate matching threads, and deliver the generated matching threads to individual cores of participating nodes; and
an aggregating unit configured to collect and sum matching results generated by the individual cores performing matching operations based on the matching threads to generate an ontology mapping.

### Claim 2 (depends on 1)

The ontology matching apparatus according to claim 1, wherein the preprocessing unit serializes the generated ontology subsets into a binary form to generate serialized ontology subsets.

### Claim 3 (depends on 2)

The ontology matching apparatus according to claim 2, further comprising an ontology storing unit configured to store the serialized ontology subsets, and provide the serialized ontology subsets that are already stored to the preprocessing unit when the same candidate ontology as the received candidate ontology is received.

### Claim 4 (depends on 3)

The ontology matching apparatus according to claim 3, wherein the preprocessing unit reconfigures the serialized ontology subsets received from the ontology storing unit through deserialization to generate ontology subsets.

### Claim 5 (depends on 1)

The ontology matching apparatus according to claim 1, wherein the matching threads include one or more matching requests corresponding to the participating nodes in one-to-one relation, one or more matching jobs corresponding to the individual cores in one-to-one relation provided in the participating nodes corresponding to the matching requests in one-to-one relation, and one or more matching tasks performing the matching operations in the individual cores.

### Claim 6 (depends on 5)

The ontology matching apparatus according to claim 5, wherein the relation among the matching requests, the matching jobs, and the matching tasks is calculated by formulae below:
MR
←
∑
i
=
1
Nodes

MR
i
MR
i
←
∑
j
=
1
Cores

MJ
j
MJ
j
←
∑
k
=
1
Copncepts

MT
k
MT
k
←
m
×
n
|
m
∈
O
s

n
∈
O
t
wherein MRi indicates the matching requests allocated to the respective participating nodes, MR indicates a set matching request of MRi allocated to the participating nodes, MJj indicates the matching jobs allocated to the individual cores provided in the participating nodes, and MTk indicates the matching tasks allocated to the individual cores to perform the matching operations.

### Claim 7 (depends on 5)

The ontology matching apparatus according to claim 5, wherein the matching jobs are independent of other matching jobs and the matching requests operating in other participating nodes.

### Claim 8 (depends on 1)

The ontology matching apparatus according to claim 1, wherein when the candidate ontologies are received, the distribution processing unit confirms the number of participating nodes available for matching and the number of individual cores provided in the participating nodes connected to the participating nodes, applies the number of the participating nodes and the number of the individual cores to a distribution algorithm to set the number of distributions, and divides the ontology subsets in consideration of the set number of distributions and applies the matching algorithm to the divided ontology subsets to generate the matching threads including matching requests, matching jobs, and matching tasks.

### Claim 9 (depends on 1)

The ontology matching apparatus according to claim 1, wherein collecting and aggregating the matching results of the aggregating unit of individual nodes including two or more participating nodes is calculated by formulae below:
O
b
Job
←
⋃
i
=
1
MT

O
b
i
,

O
b
Node
←
∑
i
=
0
MJ

O
b
job
=
i
O
b
←
∑
i
=
1
TotalNode

O
b
Node
=
i
wherein Obi indicates a matching result of a matching task (bridge ontology of the matching task), ObJob indicates an intermediate bridge ontology of a matching job that collected all results of every single matching task in all matching jobs, ObNode indicates a sum of all intermediate bridge ontologies ObJob in all matching jobs performed in each of the participating nodes, ObJob indicates intermediate bridge ontology generated in all nodes, and Ob indicates a bridge ontology that is finally generated and is converted to an ontology mapping result as an output file.

### Claim 10 (depends on 9)

The ontology matching apparatus according to claim 9, wherein the matching results calculated by all participating nodes accumulated as one ontology object referred to as a bridge ontology, and
the aggregating unit aggregates the bridge ontologies that are matching results calculated by the all participating nodes in the distributed environment and, converts the bridge ontologies to an ontology mapping result as a physical mapping file, and outputs the ontology mapping result.

### Claim 11

An ontology matching method for large-scale biomedical ontologies, the method comprising:
classifying received candidate ontologies into one or more ontology subsets;
dividing the generated ontology subsets by virtue of a distribution algorithm and applying a matching algorithm to the divided ontology subsets to generate matching threads;
delivering the generated matching threads to individual cores of participating nodes; and
collecting and accumulating matching results generated by the individual cores performing matching operations based on the matching threads to generate an ontology mapping.

### Claim 12 (depends on 11)

The ontology matching method according to claim 11, wherein the classifying of the received candidate ontologies into one or more ontology subsets includes serializing the generated ontology subsets to a binary form to generate serialized ontology subsets.

### Claim 13 (depends on 12)

The ontology matching method according to claim 12, further comprising storing the serialized ontology subsets, and reconfiguring the serialized ontology subsets that are already stored through deserialization to generate ontology subsets when the same candidate ontology as the received candidate ontology is received.

### Claim 14 (depends on 11)

The ontology matching method according to claim 11, wherein the matching threads include one or more matching requests corresponding to the participating nodes in one-to-one relation, one or more matching jobs corresponding to the individual cores in one-to-one relation provided in the participating nodes corresponding to the matching requests in one-to-one relation, and one or more matching tasks performing the matching operations in the individual cores.

### Claim 15 (depends on 14)

The ontology matching method according to claim 14, wherein the relation among the matching requests, the matching jobs, and the matching tasks is calculated by formulae below:
MR
←
∑
i
=
1
Nodes

MR
i
MR
i
←
∑
j
=
1
Cores

MJ
j
MJ
j
←
∑
k
=
1
Copncepts

MT
k
MT
k
←
m
×
n
|
m
∈
O
s

n
∈
O
t
wherein MRi indicates the matching requests allocated to the respective participating nodes, MR indicates a set matching request of MRi allocated to the participating nodes, MJj indicates the matching jobs allocated to the individual cores provided in the participating nodes, and MTk indicates the matching tasks allocated to the individual cores to perform the matching operations.

### Claim 16 (depends on 14)

The ontology matching method according to claim 14, wherein the matching jobs are independent of other matching jobs and the matching requests operating in other participating nodes.

### Claim 17 (depends on 11)

The ontology matching method according to claim 11, wherein the generating of the matching threads includes:
when the candidate ontologies are received, confirming the number of participating nodes and the number of individual cores provided in the participating nodes available in the distributed environment.
applying the number of the participating nodes and the number of the individual cores to a distribution algorithm to set the number of distributions; and
dividing the ontology subsets in consideration of the set number of distributions and applying the matching algorithm to the divided ontology subsets to generate the matching threads including matching requests, matching jobs, and matching tasks.

### Claim 18 (depends on 11)

The ontology matching method according to claim 11, wherein the collecting and summing of the matching results of the aggregating unit in the distributed environment including two or more participating nodes is calculated by formulae below:
O
b
Job
←
⋃
i
=
1
MT

O
b
i
,

O
b
Node
←
∑
i
=
0
MJ

O
b
job
=
i
O
b
←
∑
i
=
1
TotalNode

O
b
Node
=
i
wherein Obi indicates a matching result of a matching task (bridge ontology of the matching task), ObJob indicates an intermediate bridge ontology of a matching job that collected all results of every single matching task in all matching jobs, ObNode indicates a sum of all intermediate bridge ontologies ObJob in all matching jobs performed in each of the participating nodes, ObJob indicates intermediate bridge ontology generated in all nodes, and Ob indicates a bridge ontology that is finally generated and is converted to an ontology mapping result as an output file.

### Claim 19 (depends on 18)

The ontology matching method according to claim 18, wherein the collecting and accumulating of the matching results generated by the individual cores performing the matching operations based on the matching threads to generate the ontology mapping includes summing the bridge ontologies that are matching results calculated by the all participating nodes, converting the bridge ontologies to an ontology mapping result as a physical mapping file, and outputting the ontology mapping result.
