KR20130109601A - Decision method of ontology instance similarity and ontology system using the method - Google Patents

Decision method of ontology instance similarity and ontology system using the method Download PDF

Info

Publication number
KR20130109601A
KR20130109601A KR1020120031480A KR20120031480A KR20130109601A KR 20130109601 A KR20130109601 A KR 20130109601A KR 1020120031480 A KR1020120031480 A KR 1020120031480A KR 20120031480 A KR20120031480 A KR 20120031480A KR 20130109601 A KR20130109601 A KR 20130109601A
Authority
KR
South Korea
Prior art keywords
similarity
entity
class
obtaining
property
Prior art date
Application number
KR1020120031480A
Other languages
Korean (ko)
Inventor
김성혁
추윤미
김제민
박영택
Original Assignee
(주)탑쿼드란트코리아
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by (주)탑쿼드란트코리아 filed Critical (주)탑쿼드란트코리아
Priority to KR1020120031480A priority Critical patent/KR20130109601A/en
Publication of KR20130109601A publication Critical patent/KR20130109601A/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0641—Shopping interfaces
    • G06Q30/0643—Graphical representation of items or shoppers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Abstract

The present invention relates to an ontology system for determining the identity of a first entity and a second entity respectively present in two or more different ontologies according to a second aspect. The system may include: (a) a class similarity detection unit for obtaining class similarity between a first class including a first entity and a second class including a second entity; (b) an object similarity detection unit configured to determine an object similarity between the first object and the second object; an identity determining unit determining whether the first entity and the second entity are the same using the class similarity and the entity similarity; Respectively.

Description

Decision method of Ontology instance similarity and Ontology system using the method

The present invention relates to an ontology system based on two or more different ontologies, and more particularly, in linking two or more different ontologies, the same entity existing in different ontologies automatically into one entity. It relates to a method for determining ontologies of individual similarity.

Ontology-based services are often built on two or more different ontologies. Ontologies have different URIs to identify their resources. Therefore, in the real world, even the same entity is recognized as a different entity between different ontologies. 1 is a diagram illustrating an entity existing in two different ontologies. Referring to FIG. 1, the mo: ICJ object of the mo: MusicArtist class in Music Ontology and the lmr: ICJ object of the lmc: Person class in Movie Ontology refer to singer and actor "ICJ". . However, since the two ontology assign different URIs mo and lmr to ICJ entities, it is difficult to construct a service system that links music information and movie information based on music and film ontology. In other words, the music ontology has no softening information appeared by the ICJ, and the movie ontology has no record information published by the ICJ.

The most important point in linking two or more ontologies is to distinguish between identical or semantically equivalent entities in the real world, which is impossible to do manually because of the vast amount of individual information.

Therefore, algorithms for automatically classifying semantically equivalent entities in two or more different ontologies into one entity have been proposed.

The above algorithm has a method of applying the hasKey axiom of OWL2RL. OWL2 is a standard ontology authoring language classified into OWL2EL, OWL2RL, and OWL2QL according to its expression power. Among them, OWL2RL provides axioms (formalized rules) for standard ontology inference. Therefore, if a specific knowledge area is defined as an OWL2RL-based ontology, various knowledge can be expanded through the inference engine.

The hasKey axiom in OWL2RL states, “A class can specify some of its attributes as key attributes. And for objects of this class, all property values specified as key properties concatenate the same objects semantically equal. ”

FIG. 2 is a diagram illustrating the hasKey axiom of OWL2RL, which is a standard ontology authoring language, and its meaning. Referring to FIG. 2, in order to connect the objects of two ontologies by applying haskey axiom, first, the class hierarchy and property relations of the two ontologies must be changed. Normally, this process is not very heavy because it is performed on ontology schema. Therefore, there is a problem that a person must perform directly based on expert knowledge. Automatic research is also underway at many research institutes, but with less accuracy.

FIG. 3 exemplarily illustrates the concatenation of the same entities in the music ontology and the movie ontology in a conventional manner.

Referring to FIG. 3, in order to connect the semantically identical objects existing in the music ontology and the movie ontology into one, the mo: MusicArtist class of the music ontology is subordinate to the lmc: Person class of the film ontology. Set to class. In this way, an instance of mo: MusicArtist is inferred as an object of class lmc: Person by an instance axiom in OWL2RL. Next, set the km: realName and km: birthday attributes of the mo: MusicArtist class and the lmp: name and lmp: birthday of the lmc: Person class as equivalent properties. Finally, set km: realName and km: birthday as key attributes.

After this preliminary work is completed, objects in the mo: MusicArtist class that have the same km: realName and lmp: name, km: birthday, and lmp: birthday values are connected to one object in the lmc: Person class.

However, the above-described conventional method has a problem in that when a key attribute is not specified or one attribute value designated as a key attribute is missing, it is not connected to one entity.

(1) Korean Patent Publication No. 10-2007-61353 (2) Korean Patent Publication No. 10-2008-45823

An object of the present invention for solving the above problems is to provide an ontology entity similarity determination method for automatically determining the identity of the entities belonging to two different ontologies and an ontology system using the same.

A first aspect of the present invention for achieving the above technical problem relates to a method for determining the identity of the first entity and the second entity respectively present in two or more different ontology, (a) the first entity (i 1 Obtaining a class similarity (Class Distance (C 1 , C 2 )) of the first class C 1 including ) and the second class C 2 including the second entity i 2 ; (b) obtaining an entity similarity (Rdistance (i 1 , i 2 )) of the first entity and the second entity; (c) determining whether the first entity and the second entity are the same using the class similarity and the entity similarity; It is provided.

In the method according to the first aspect described above, the step (c) comprises: (c1) obtaining an overall similarity value (SI) of the first entity and the second entity using the class similarity and the entity similarity; (c2) determining that the first entity and the second entity are the same when the total similarity value is equal to or greater than a preset threshold.

In the method according to the first aspect described above, obtaining the Class Similarity (Class Distance (C 1 , C 2 )) of the step (a) comprises: (a1) a first set of the upper class of the first entity and the first; Obtaining a second set of higher classes of the two entities; (a2) obtaining a first number which is the number of upper classes included in the first set and the second set in common; (a3) obtaining a second number which is the total number of upper classes included in the first set and the second set; (a4) setting a value obtained by dividing the first number by the second number to a class similarity (Class Distance (C 1 , C 2 )).

In the method according to the first feature described above, obtaining the object similarity (Rdistance (i 1 , i 2 )) of the step (b), (b1) the data property similarity (rdist) of the first object and the second object obtaining ((i 1 , i 2 )); (b2) obtaining object property similarity (ordist ((i 1 , i 2 )) of the first object and the second object; and (b3) the data property similarity And obtaining the object similarity (Rdistance (i 1 , i 2 )) using the object property similarity.

The step of obtaining a data attribute similarity rdist ((i 1 , i 2 )) of the first entity and the second entity may include setting the data attribute similarity at the start of determining similarity between the first entity and the second entity. After initializing to, extracting common data attributes of the first entity and the second entity among the data attributes of the first entity and the second entity, and if the common data attributes have the same value, the data attribute similarity is' 1. 'Increasing by,

The first object and the step of obtaining an object attribute similarity of two objects (ordist ((i 1, i 2)), the object type attribute value of the first object (ori 1) and an object type attribute value of the second object ( ori 2) object similarity (Rdistance (ori 1, ori 2 )) to obtain a first class, belonging to the object type attribute value of the object (C ori1) and the second class belongs to an object type attribute value of the object for a (C ori2 Class distance similarity (C ori1 , C ori2 )), and object property similarity (ordist ((i 1 , i 2 )) by using the object similarity and the class similarity.

The object similarity (Rdistance) is determined as a value obtained by dividing the sum of the data property similarity and the object property similarity between the first object and the second object by the number of properties (n), and the number of properties (n) is the first and second objects. It is characterized in that the number of properties having the class to which the object belongs.

An ontology system for determining the identity of a first entity and a second entity respectively present in two or more different ontologies according to the second aspect of the present invention, includes: (a) a first class and a second entity including the first entity A class similarity detector for obtaining class similarity with respect to the second class included; (b) an object similarity detection unit configured to determine an object similarity between the first object and the second object; an identity determining unit determining whether the first entity and the second entity are the same using the class similarity and the entity similarity; Respectively.

In the ontology system according to the second aspect described above, the identity determination unit obtains the total similarity value (SI) of the first object and the second object using the class similarity and the object similarity, and the total similarity value is preset. If it is more than the threshold, it is preferable to determine that the first entity and the second entity are the same.

In the ontology system according to the second aspect described above, the class similarity detection unit obtains a first set of an upper class of a first entity and a second set of an upper class of a second entity and is common to the first set and the second set. A first number, which is the number of upper classes included in, is obtained, a second number, which is the number of all upper classes included in the first set and the second set, and a value obtained by dividing the first number by a second number is class similarity. It is preferable to set to.

In the ontology system according to the second aspect described above, the object similarity detecting unit includes: (b1) a data attribute similarity detecting unit for obtaining data attribute similarities between the first object and the second individual; (b2) an object property similarity detector for obtaining object property similarity between the first object and the second object; And (b3) an entity similarity determining unit that obtains individual similarity using the data attribute similarity and the object attribute similarity.

In the ontology system according to the second aspect described above, the data attribute similarity detection unit initializes the data attribute similarity to a preset value when starting the similarity determination between the first entity and the second entity, and then initializes the data attribute similarity with the first entity. And extracting common data attributes of the first entity and the second entity among the data attributes of the second entity, and increasing the data attribute similarity by '1' when the common data attributes have the same value.

The object property similarity is obtained from the object similarity value between the object type property value of the first object and the object type property value of the second object, and the object type property value of the first object and the object type property value of the second object. It is characterized by obtaining the class similarity for the class to which it belongs, and the object property similarity using the object similarity and the class similarity,

The entity similarity is determined by dividing the sum of the data attribute similarity and the object attribute similarity between the first entity and the second entity by the number of attributes n, and the number of attributes n corresponds to the first and second entities. It is characterized in that the number of properties having a given class in the airspace.

By the ontology entity similarity determination method and system according to the present invention, it is possible to automatically distinguish the same entity into a single entity in the meaning of two or more different ontologies. In particular, according to the present invention, even if the ontology does not have a key attribute specified or if any attribute value designated as a key attribute is missing, even if the entities belong to different ontologies, semantically identical entities can be linked to one entity. Will be.

1 is a diagram illustrating an entity existing in two different ontologies.
FIG. 2 is a diagram illustrating the hasKey axiom of OWL2RL, which is a standard ontology authoring language, and its meaning.
FIG. 3 exemplarily illustrates the concatenation of the same entities in the music ontology and the movie ontology in a conventional manner.
4 is a block diagram schematically illustrating a configuration of an ontology system according to a preferred embodiment of the present invention.
5 is an ontology schema diagram illustrating an example of a class similarity detection unit calculating a class similarity of an ontology system according to an exemplary embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings will be described in detail with respect to the ontology system for determining the identity of the first entity and the second entity respectively present in two or more different ontologies according to a preferred embodiment of the present invention. In particular, the present invention is characterized by providing a method for solving this problem when the hasKey axiom of OWL2RL is not applied or the ontology object cannot be determined by hasKey axiom alone.

4 is a block diagram schematically illustrating a configuration of an ontology system according to a preferred embodiment of the present invention. Referring to FIG. 4, the ontology system according to the present invention includes a class similarity detection unit 100, an object similarity detection unit 110, and an identity determination unit 120, and each of the first entities existing in two different ontologies. It is determined whether and the second entity is the same and output the result. Hereinafter, operations of the above-described elements will be described in detail.

The class likelihood detection unit 100 includes a first object (i1) belongs to a first class (C 1) and a second object (i 2) the class similarity for the second class belongs (C 2) (Class Distance (C 1 , C 2 )). An ontology object always belongs to a class that reflects its characteristics, and each class is divided into subclasses according to the difference in the detailed characteristics. Thus, the first criterion for determining that the ontology's objects are equal is the Class Similarity (CD) of the objects. The class similarity detector detects class similarity (Class distance (i 1 , i 2 )) of the first entity and the second entity using Equation 1.

Figure pat00001

Here, C 1 is a class of the first entity, H C1 is a set of superclasses of C 1 , C 2 is a class of second entities, and H C2 is a set of superclasses of C 2 . Accordingly, class similarity (Class distance (i 1 , i 2 )) according to Equation 1 is obtained by first obtaining a first set of an upper class of a first entity and a second set of an upper class of a second entity. Obtaining a first number, which is the number of higher classes commonly included in a second set, obtaining a second number, which is the total number of higher classes included in the first set and the second set, and dividing the first number by a second number Set the value to class similarity (Class distance (i 1 , i 2 )). This class similarity makes it possible to determine how much different first and second entities share characteristics.

Hereinafter, a process of calculating class similarity will be described with reference to FIG. 5. 5 is an ontology schema diagram illustrating an example of a class similarity detection unit calculating a class similarity of an ontology system according to an exemplary embodiment of the present invention.

C 1 , C 2 , (C 1 , H C1 ), (C 2 , H C2 ), | (C 1 , H C1 ) ∪ (C 2 , H C2 ) |, | (C in the ontology in FIG. 5 1 , H C1 ) ∩ (C 2 , H C2 ) | And Class Distance (C 1 , C 2 ) is the same as Equation 2.

Figure pat00002

On the other hand, each entity of the ontology has a value for an object type attribute and a data type attribute representing its characteristics. For example, an entity belonging to a person class would have a value for a data type attribute such as name, date of birth, gender, etc., and an object type attribute that describes the relationship between another entity. Thus, the second criterion for determining the identity of an entity of an ontology is the data attribute similarity and object attribute similarity for the entities.

The object similarity detecting unit 110 includes a data attribute similarity detecting unit 112, an object attribute similarity detecting unit 114, and an object similarity determining unit 116, so that data attribute similarity (rdist) and object attribute similarity (ordist) between objects. ) And use it to determine individual similarity (Rdistance).

The data attribute similarity detection unit 112 obtains the data attribute similarity rdist (i 1 , i 2 ) of the first entity and the second entity, and the data attribute at the beginning of the similarity determination of the first entity and the second entity. After the similarity is initialized to a preset value, the common data attributes p of the first entity and the second entity among the data attributes of the first entity and the second entity are extracted, and the common data attributes are set to the same value. If so, the data attribute similarity is increased by '1'. The similarity of data attributes of the first entity i 1 and the second entity i 2 can be obtained by equation (3).

Figure pat00003

Where p is a common data attribute of the first entity and the second entity, As (p, i 1 ) is the value of the data attribute p for the first entity, and As (p, i 1 ) = 0 is zero 1 means there is no value of the property (p) of the object. Therefore, when there is an attribute having the same value among attributes that the first entity and the second entity have in common, the data attribute similarity rdist (i 1 , i 2 ) is increased by one.

The object property similarity detection unit 114 obtains the object property similarity (ordist (i 1 , i 2 )) of the first object and the second object, and the object property similarity (ordist (i 1 ) of the first object and the second object. , i 2 )) can be obtained using Equation 4.

Figure pat00004

Referring to Equation 4, the object property similarity (ordist (i 1 , i 2 )) of the first object and the second object is the object type property value of the first object (ori 1 ) and the object type property value of the second object ( ori 2 ), the object similarity (Rdistance (ori 1 , ori 2 )), and the class (C ori1 ) to which the object type property value of the first object (ori 1 ) belongs and the object type property value of the second object (ori 2). ) is obtained by using a (class similarity (class Distance (C ori1, C ori2 for C ori2))) belonging to this class.

That is, the object property similarity (ordist (i 1 , i 2 )) of the first object and the second object is (1) the object type property value of the first object (ori 1 ) and the object type property value of the second object (ori the object similarity to 2) (Rdistance (ori 1, ori 2)) to obtain and, (2) first object type attribute value of the object (ori class to which it belongs 1) (C ori1) with an object type attribute value of the second object, obtains the class similarity (Class Distance (C ori1 , C ori2 )) of the class (C ori2 ) to which (ori 2 ) belongs, and (3) the object type property value (ori 1 ) of the first object and the object of the second object. The object similarity (Rdistance (ori 1 , ori 2 )) to the type property value (ori 2 ), and the object type property of the class (C ori1 ) to which the object type property value (ori 1 ) of the first object belongs and the second object. Finally, the object attribute similarity (ordist (i 1 , i 2 )) is obtained by obtaining an average value of the class similarity (Class Distance (C ori1 , C ori2 )) of the class C ori2 to which the value (ori 2 ) belongs.

Accordingly, the object property similarity (ordist (i 1 , i 2 )) of the first and second objects is a combination of the property similarity between the property similarities between the property values of the first and second objects in common. It is composed.

The entity similarity determining unit 116 determines the entity similarity (Rdistance (i 1 , i 2 )) of the first entity and the second entity using the data attribute similarity and the object attribute similarity. The object similarity (Rdistance (i 1 , i 2 )) of the first object and the second object is the data property similarity (rdist (i 1 , i 2 )) and the object property similarity (ordist) of the first object and the second object. The sum of (i 1 , i 2 )) is divided by the number n of attributes, and the number n of attributes is the number of attributes having a class to which the first and second entities belong. Equation 5 is an equation expressing the object similarity (Rdistance (i 1 , i 2 )) of the first entity and the second entity.

Figure pat00005

When the ontology and the entity according to FIG. 5 exist, a calculation process for obtaining the entity similarity between the first entity preview001 and the second entity preview002 is shown in Equation 6 below. Through Equation 6, the object similarity Rdistance (i 1 , i 2 ) is 0.313.

Figure pat00006

The identity determination unit 120 determines whether the first entity and the second entity are the same by using the class similarity and the entity similarity.

The identity determining unit may determine an overall similarity value SI between the first entity and the second entity by using the class distance (C 1 , C 2 ) and the entity similarity (Rdistance (i 1 , i 2 )). If the total similarity value is equal to or greater than a preset threshold value θ, the first entity and the second entity are determined to be the same. Equation 7 shows a process of obtaining the overall similarity SI of the first and second entities.

Figure pat00007

The threshold value θ is a value adjusted appropriately by the developer according to the ontology-based system to be built. In Equation 7, the output value is a Boolean value indicating whether the first and second entities are equal. Through Equation 7, a value indicating whether the first and second entities are the same is finally output.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be understood that various changes and modifications may be made without departing from the spirit and scope of the invention. It is to be understood that the present invention may be embodied in many other specific forms without departing from the spirit or essential characteristics thereof.

Ontology individual similarity determination method according to the present invention can determine the identity between the entities belonging to different ontologies to automatically classify as one entity, so when constructing a system based on two or more different ontology It can be widely used.

100: class similarity detection unit
110: object similarity detection unit
120: identity determination unit
112: data attribute similarity detection unit
114: object property similarity detection unit
116: object similarity determining unit

Claims (14)

In the method of determining the identity of the first individual and the second individual respectively present in two or more different ontologies,
(a) the first entity (i 1) a first class (C 1) and a second object class similarity of the second class (C 2) contains an (i 2) which contains a (Class Distance (C 1, C 2, Obtaining));
(b) obtaining an entity similarity (Rdistance (i 1 , i 2 )) of the first entity and the second entity;
(c) determining whether the first entity and the second entity are the same using the class similarity and the entity similarity;
Ontology individual similarity determination method comprising the. 2. The method of claim 1, wherein step (c)
(c1) obtaining an overall similarity value (SI) of the first entity and the second entity using the class similarity and the entity similarity;
(c2) determining that the first entity and the second entity are the same when the total similarity value is greater than or equal to a preset threshold;
Ontology individual similarity determination method comprising the. The method of claim 1, wherein the obtaining of the class similarity (Class Distance (C 1 , C 2 )) of the step (a) is
(a1) obtaining a first set of higher classes of the first entity and a second set of higher classes of the second entity;
(a2) obtaining a first number which is the number of upper classes included in the first set and the second set in common;
(a3) obtaining a second number which is the total number of upper classes included in the first set and the second set;
(a4) setting a value obtained by dividing the first number by a second number to a class similarity (Class Distance (C 1 , C 2 ));
Ontology individual similarity determination method comprising the. The method of claim 1, wherein the obtaining the individual similarity (Rdistance (i 1 , i 2 )) of the step (b),
(b1) obtaining a data attribute similarity rdist ((i 1 , i 2 )) of the first entity and the second entity;
(b2) obtaining object property similarity ordist ((i 1 , i 2 )) of the first entity and the second entity; and
(b3) obtaining an object similarity (Rdistance (i 1 , i 2 )) using the data property similarity and the object property similarity;
Ontology individual similarity determination method comprising the. The method of claim 4, wherein obtaining the data property similarity rdist ((i 1 , i 2 )) of the first entity and the second entity is as follows.
After initializing the similarity of the data attributes to a preset value at the beginning of the determination of the similarity between the first entity and the second entity, common data of the first entity and the second entity among the data attributes of the first entity and the second entity And extracting the attributes and increasing the data attribute similarity by '1' when the common data attributes have the same value. The method of claim 4, wherein the obtaining of the object property similarity (ordist ((i 1 , i 2 )) of the first object and the second object is performed by:
Obtain the object similarity (Rdistance (ori 1 , ori 2 )) between the object type property value (ori 1 ) of the first object and the object type property value (ori 2 ) of the second object, and obtain the object type property value of the first object. Obtain a class similarity (Class distance (C ori1 , C ori2 )) for the class (C ori1 ) to which it belongs and the object type attribute value of the second object (C ori2 ), and use the object similarity and the class similarity Ontology object similarity determination method characterized in that to obtain the object property similarity (ordist ((i 1 , i 2 )). The method of claim 4, wherein the object similarity (Rdistance) is determined as a value obtained by dividing the sum of data property similarity and object property similarity between the first object and the second object by the number of properties (n), and the number of properties (n). The ontology individual similarity determination method, characterized in that the number of properties having the class belonging to the first and second entities in the airspace. In the ontology system for determining the identity of the first entity and the second entity respectively present in two or more different ontologies,
(a) a class similarity detector for obtaining class similarity between a first class including a first entity and a second class including a second entity;
(b) an object similarity detection unit configured to determine an object similarity between the first object and the second object;
an identity determining unit determining whether the first entity and the second entity are the same using the class similarity and the entity similarity;
Ontology system comprising a. The method of claim 8, wherein the identity determination unit
The total similarity value SI of the first entity and the second entity is obtained using the class similarity and the entity similarity, and if the total similarity value is equal to or greater than a preset threshold, the first entity and the second entity are determined to be the same. Ontology system, characterized in that. The method of claim 8, wherein the class similarity detection unit
Obtaining a first set of upper classes of the first entity and a second set of upper classes of the second entity, obtaining a first number which is the number of upper classes commonly included in the first set and the second set, and An ontology system, comprising: obtaining a second number, which is the number of all upper classes included in the second set, and setting a value obtained by dividing the first number by a second number as a class similarity. The method of claim 8, wherein the individual similarity detection unit,
(b1) a data property similarity detector for obtaining data property similarity between the first entity and the second entity;
(b2) an object property similarity detector for obtaining object property similarity between the first object and the second object; And
(b3) an entity similarity calculator which calculates entity similarity using the data attribute similarity and the object attribute similarity;
Ontology system comprising a. The method of claim 11, wherein the data attribute similarity detection unit,
After initializing the similarity of the data attributes to a preset value at the beginning of the determination of the similarity between the first entity and the second entity, common data of the first entity and the second entity among the data attributes of the first entity and the second entity And extract the attributes and increase the data attribute similarity by '1' when the common data attributes have the same value. The method of claim 11, wherein the object property similarity is
Obtain the object similarity between the object type property value of the first object and the object type property value of the second object, and the class to which the object type property value of the first object belongs and the class to which the object type property value of the second object belongs. Obtaining similarity, and using the object similarity and the class similarity to obtain the object property similarity ontology system. The method of claim 11, wherein the object similarity is determined by dividing the sum of data property similarity and object property similarity between the first object and the second object by the number of properties n, and the number of properties n is the first value. And the number of attributes having the class to which the second entity belongs as an airspace.

KR1020120031480A 2012-03-28 2012-03-28 Decision method of ontology instance similarity and ontology system using the method KR20130109601A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020120031480A KR20130109601A (en) 2012-03-28 2012-03-28 Decision method of ontology instance similarity and ontology system using the method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020120031480A KR20130109601A (en) 2012-03-28 2012-03-28 Decision method of ontology instance similarity and ontology system using the method

Publications (1)

Publication Number Publication Date
KR20130109601A true KR20130109601A (en) 2013-10-08

Family

ID=49631796

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020120031480A KR20130109601A (en) 2012-03-28 2012-03-28 Decision method of ontology instance similarity and ontology system using the method

Country Status (1)

Country Link
KR (1) KR20130109601A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108021703A (en) * 2017-12-26 2018-05-11 广西师范大学 A kind of talk formula intelligent tutoring system
CN110427606A (en) * 2019-06-06 2019-11-08 福建奇点时空数字科技有限公司 A kind of professional entity similarity calculating method based on semantic model
CN113535968A (en) * 2020-04-20 2021-10-22 北京沃东天骏信息技术有限公司 Method and device for extracting key attributes of data

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108021703A (en) * 2017-12-26 2018-05-11 广西师范大学 A kind of talk formula intelligent tutoring system
CN110427606A (en) * 2019-06-06 2019-11-08 福建奇点时空数字科技有限公司 A kind of professional entity similarity calculating method based on semantic model
CN113535968A (en) * 2020-04-20 2021-10-22 北京沃东天骏信息技术有限公司 Method and device for extracting key attributes of data

Similar Documents

Publication Publication Date Title
US9542477B2 (en) Method of automated discovery of topics relatedness
CA3007723C (en) Systems and/or methods for automatically classifying and enriching data records imported from big data and/or other sources to help ensure data integrity and consistency
RU2619193C1 (en) Multi stage recognition of the represent essentials in texts on the natural language on the basis of morphological and semantic signs
US10740380B2 (en) Incremental discovery of salient topics during customer interaction
Srivastava et al. A novel hierarchical BERT architecture for sarcasm detection
US10198432B2 (en) Aspect-based sentiment analysis and report generation using machine learning methods
US20130246321A1 (en) Refinement and calibration mechanism for improving classification of information assets
US9262506B2 (en) Generating mappings between a plurality of taxonomies
CN103488671A (en) Method and system for querying and integrating structured and instructured data
CN111931049B (en) Business processing method based on big data and artificial intelligence and block chain financial system
Iqbal et al. Balancing prediction errors for robust sentiment classification
WO2022042297A1 (en) Text clustering method, apparatus, electronic device, and storage medium
KR20160104064A (en) A multidimensional recursive learning process and system used to discover complex dyadic or multiple counterparty relationships
Giancristofaro et al. Predicting sentiment toward transportation in social media using visual and textual features
Sapul et al. Trending topic discovery of Twitter Tweets using clustering and topic modeling algorithms
CN109783633A (en) Data analysis service procedural model recommended method
CN115146062A (en) Intelligent event analysis method and system fusing expert recommendation and text clustering
Zhang et al. Development of a radiology decision support system for the classification of MRI brain scans
Baby et al. An efficient lymphocytic leukemia detection based on EfficientNets and ensemble voting classifier
KR20130109601A (en) Decision method of ontology instance similarity and ontology system using the method
RU2681356C1 (en) Classifier training used for extracting information from texts in natural language
US20190122232A1 (en) Systems and methods for improving classifier accuracy
US20220261863A1 (en) Method and system for transitory sentiment community- based digital asset valuation
WO2023178767A1 (en) Enterprise risk detection method and apparatus based on enterprise credit investigation big data knowledge graph
Acosta-Mendoza et al. A new algorithm for approximate pattern mining in multi-graph collections

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E601 Decision to refuse application