Calculating term similarity using a meta-model semantic network

Patent metadata

Field	Value
Publication number	US-9256595-B2
Application number	US-201113284695-A
Country	US
Kind code	B2
Filing date	Oct 28, 2011
Priority date	Oct 28, 2011
Publication date	Feb 9, 2016
Grant date	Feb 9, 2016

Abstract

In example embodiments, a technique is provided to determine the similarity between two terms. For example, example embodiments may store a meta-model semantic network that includes a first and second term. Further, both the first and second terms are respectively associated with model and meta-model information. A request to calculate a term similarity value is received. A term similarity value expresses a correlation between the first term and the second term. The term similarity value is then calculated based on a comparison of the model and the meta-model information associated with the first and second terms.

First claim

What is claimed is:

1. A method comprising:
storing a meta-model semantic network, the meta-model semantic network including a first term and a second term, the first term being associated with a first model and a first meta-model, the second term being associated with a second model and a second meta-model, wherein the first model comprises one or more first attributes that characterize the first term, wherein the second model comprises one or more second attributes that characterize the second term, wherein the first meta-model characterizes the first model, wherein the second meta-model characterizes the second model, and wherein a link and a link attribute define a relationship between at least two nodes of the semantic network;
receiving a request to calculate a term similarity value, the term similarity value expressing a correlation between the first term and the second term;
calculating, using at least one processor, the term similarity value based on a distance between one of the first attributes of the first model and one of the second attribute of the second model, a distance between an attribute of the first meta-model and an attribute of the second meta-model, and a distance between an attribute of the first term and an attribute of the second term; and
modifying one or more of a search query or the meta-model semantic network based on the term similarity value.

2. The method of claim 1 , further comprising:
linking the first term to a concept in the meta-model semantic network; and
based on the calculated term similarity value, linking the second term to the concept.

3. The method of claim 1 , further comprising defining a first term sense associated with the first term, the first term sense including specified models and specified meta-models of the meta-model semantic network.

4. The method of claim 3 , wherein the calculation of the term similarity value comprises selecting the first model and the first meta-model based on the specified models and the specified meta-models included in the first term sense.

5. The method of claim 1 , further comprising sending an indication that the first term and the second term in response to the similarity value meeting a threshold value.

6. The method of claim 1 , wherein the calculation of the term similarity value comprises a comparison of:
a first link associated with the first term and a second link associated with the second term;
a first link attribute associated with the first link and a second link attribute associated with the second link; and
a first link attribute definition associated with the first link attribute and a second link attribute definition associated with the second link attribute definition.

7. A non-transitory, machine-readable medium that stores instructions, which, when performed by a machine, cause the machine to perform operations comprising:
storing a meta-model semantic network, the meta-model semantic network including a first term and a second term, the first term being associated with a first model and a first meta-model, the second term being associated with a second model and a second meta-model, wherein the first model comprises one or more first attributes that characterize the first term, wherein the second model comprises one or more second attributes that characterize the second term, wherein the first meta-model characterizes the first model, wherein the second meta-model characterizes the second model, and wherein a link and a link attribute define a relationship between at least two nodes of the semantic network;
receiving a request to calculate a term similarity value, the term similarity value expressing a correlation between the first term and the second term; and
calculating the term similarity value based on a distance between one of the first attributes of the first model and one of the second attributes of the second model, a distance between an attribute of the first meta-model and an attribute of the second meta-model, and a distance between an attribute of the first term and an attribute of the second term; and
modifying one or more of a search query or the meta-model semantic network based on the term similarity value.

8. The non-transitory, machine-readable medium of claim 7 , further comprising:
linking the first term to a concept in the meta-model semantic network; and
based on the calculated term similarity value, linking the second term to the concept.

9. The non-transitory, machine-readable medium of claim 7 , further comprising defining a first term sense associated with the first term, the first term sense including specified models and specified meta-models of the meta-model semantic network.

10. The non-transitory, machine-readable medium of claim 9 , wherein the calculation of the term similarity value comprises selecting the first model and the first meta-model based on the specified models and the specified meta-models included in the first term sense.

11. The non-transitory, machine-readable medium of claim 7 , further comprising sending an indication that the first term and the second term are similar in response to the similarity value meeting a threshold value.

12. The non-transitory, machine-readable medium of claim 7 , wherein the calculation of the term similarity value comprises a comparison of:
a first link associated with the first term and a second link associated with the second term;
a first link attribute associated with the first link and a second link attribute associated with the second link; and
a first link attribute definition associated with the first link attribute and a second link attribute definition associated with the second link attribute definition.

13. A system comprising:
a memory storage system to store a meta-model semantic network, the meta-model semantic network including a first term and a second term, the first term being associated with a first model and a first meta-model, the second term being associated with a second model and a second meta-model, wherein the first model comprises one or more first attributes that characterize the first term, wherein the second model comprises one or more second attributes that characterize the second term, wherein the first meta-model characterizes the first model, wherein the second meta-model characterizes the second model, and wherein a link and a link attribute define a relationship between at least two nodes of the semantic network;
a computer system in communication with the memory storage system, wherein the computer system comprises at least one processor and a computer-readable storage medium coupled to the at least one processor, the computer-readable storage medium comprising code executable by the processor for implementing a method comprising:
accessing the meta-model semantic network;
receiving a request to calculate a term similarity value, the term similarity value expressing a correlation between the first term and the second term;
calculating the term similarity value based on a distance between one of the first attributes of the first model and one of the second attributes of the second model, a distance between an attribute of the first meta-model and an attribute of the second meta-model, and a distance between an attribute of the first term and an attribute of the second term; and
modifying one or more of a search query or the meta-model semantic network based on the term similarity value.

14. The system of claim 13 , wherein the method further comprises:
linking the first term to a concept in the meta-model semantic network; and
based on the calculated term similarity value, linking the secon

Assignees

• Heidasch Robert
• Scheidl Stefan
• Neumann Michael
• Brand Stephan
• Licht Nico
• Reichenberger Klaus
• Heimann Archim
• Moldaner Steffen
• Sap Se
• Intelligent Views Gmbh

Inventors

• Heidasch Robert
• Scheidl Stefan
• Neumann Michael
• Brand Stephan
• Licht Nico
• Reichenberger Klaus
• Heimann Archim
• Moldaner Steffen

Classifications

• G06F40/247 Primary

  Thesauruses; Synonyms · CPC title

• G06F16/367 Primary

  Ontology · CPC title

• G06N5/02

  Knowledge representation; Symbolic representation · CPC title

• G06F40/30

  Semantic analysis · CPC title

• G06F17/2795 Primary

  Physics · mapped topic