CN107766556A - A kind of interactive Ontology Matching method and computer equipment based on evolution algorithm - Google Patents
A kind of interactive Ontology Matching method and computer equipment based on evolution algorithm Download PDFInfo
- Publication number
- CN107766556A CN107766556A CN201711070247.0A CN201711070247A CN107766556A CN 107766556 A CN107766556 A CN 107766556A CN 201711070247 A CN201711070247 A CN 201711070247A CN 107766556 A CN107766556 A CN 107766556A
- Authority
- CN
- China
- Prior art keywords
- mrow
- matching
- ontology
- result
- msub
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
The present invention provides a kind of interactive Ontology Matching method based on evolution algorithm, including body division stage, large-scale body is divided into small-scale body piecemeal so that follow-up Ontology Matching process is carried out in body piecemeal;In the Ontology Matching stage based on evolution algorithm, the Ontology Matching process of automation is realized using evolution algorithm, adaptively determine the time point of user's intervention;Subscriber checking and assay diffusion phase, the candidate's Ontology Matching result automatically determined is presented to subscriber checking, remaining candidate's Ontology Matching result is automatically processed during subscriber checking, finally spreads believable subscriber checking result;Body divided-fit surface result integrates and evaluation phase, integrates different body divided-fit surface results, and the f measurements of the reference matching result calculating final result using body.
Description
Technical field
The present invention relates to the computer equipment of a kind of Ontology Matching method and executable this method.
Background technology
Body:The clear and definite specification explanation of generalities to the knowledge in some field, i.e., to right present in some field
As, the formal and standardization description of concept, other entities and the relation between them.
Ontology Matching process:Determine that semantic identical entity is to the process of set in two heterogeneous ontologies.
Ontology Matching technology can solve the problems, such as bulk heteroj (same concept has different names in different bodies), real
The now cooperation based on the application program of body on semantic level.
Core technology of the body as semantic net, it is that a kind of shared, formal information exchanges reference model, it is described
Object present in some field, concept, other entities and the relation [1] between them.Ontology is in information management, letter
The fields such as breath retrieval, ecommerce and biomedicine all have important application, contain in the body announced at present in semantic net
Entity scale reached 3,100,000,000.However, (same concept has not the heterogeneous problem between these bodies in different bodies
Same name) it is the biggest obstacle for realizing semantic coordination between different application in semantic net, and restrict the bottleneck of semantic net development
Problem.Ontology Matching technology is that currently solve the problems, such as bulk heteroj most efficient method [2].It is but complete by this patent return
Into (2 months 2017) before, existing Ontology Matching technology has close to the 500000000 Entities Matching results announced in semantic net
45% is wrong [3,4].Therefore, it is badly in need of a kind of technology that can obtain high quality Ontology Matching result in the industry.Due to body
The complexity of matching process, the Ontology Matching result of automation Ontology Matching technical limit spacing need to ensure it by user's checking
Quality.User and automation Ontology Matching technology is made to cooperate with each other within reasonable time to obtain the Ontology Matching knot of high quality
The process of fruit is referred to as interactive Ontology Matching process.Body matching system can not be required in one Ontology Matching result of user's checking
Whole Entities Matchings pair, therefore one of existing significant challenge problem is how to minimize during interactive Ontology Matching
The value of subscriber checking result is maximized while the workload of user.
In the work for minimizing amount of user effort, Shi etc. [5] proposes the threshold value and similar by interactive remote teaching determination
Scatter diagram is spent to select most to have the matching of the problem of information content to having the user verify that.[6] such as Jim é nez-Ruiz propose three originals
Then (i.e. uniformity, locality and retention principle) filters candidate's Entities Matching pair.Beisswanger etc. [7] proposes one
A little Quality Inspection Criterias measure the reusability of Ontology Matching result, and use its determination candidate's Entities Matching collection.Cruz[8]
Deng selection allow different Ontology Matching devices can not reach an agreement result the problem of Entities Matching to user.Similarly, SAMBO
[9] knowledge crossed by user's checking before reduces unnecessary user intervention.GOMMA [10] using based on combination and
The adaptive algorithm of otherness is multiplexed impregnable Entities Matching pair.The knot that PROMPT [11] passes through newest user intervention
Fruit determines candidate's Entities Matching collection.
It is general that the result of user's checking can be diffused into its neighbour by similarity broadcast algorithm according to Ontological concept architecture
It is the effective ways for maximizing user's checking result in thought.Shi etc. [5] proposes a kind of Active Learning framework, can to
Family offer most has the candidate matches result of information content to user's checking, and by user's checking result according to Ontological concept architecture
Spread to improve the accuracy of matching.AgreementMaker [8] is diffused into using the feedback result of user for signature vectors
Other Entities Matching centerings are gone, and are realized by a linear function to increase and reduction of the relevant matches to Similarity value.
Sum up these existing Ontology Matching methods at present, following shortcomings be present:
(1) due to each iterative process will user intervention, add the unnecessary workload of user;
(2) need user to verify candidate matches result one by one, can not be automatically processed according to the result of user remaining
Candidate matches result, add the workload of user;
(3) the user's checking result of easy spreading errors, so as to reduce the quality of Ontology Matching.
[1]Garrido A.Logical Foundations of Artificial Intelligence[J]
.BRAIN.Broad Research in Artificial Intelligence andNeuroscience,2010,1(2):
149-152.
[2]Shvaiko P,Euzenat J.Ontology matching:state of the art and future
challenges[J].IEEE Transactions onknowledge anddataengineering,2013,25(1):
158-176.
[3]Liu W.Truth discovery to resolve object conflicts in linked data
[J].arXiv preprint arXiv:1509.00104,2015.
[4]Liu W,Liu J,Duan H,et al.Exploiting Source-Object Network to
Resolve Object Conflicts in LinkedData[J].arXivpreprint arXiv:1604.08407,
2016.
[5]F.Shi,J.Li,J.Tang,G.Xie,and H.Li,“Actively learning ontology
matching via user interaction,”in International Semantic Web
Conference.Springer,2009,pp.585–600.
[6]E.Jiménez-Ruiz,B.C.Grau,I.Horrocks,and R.Berlanga,“Logic-based
assessment ofthe compatibility ofumls ontology sources,”Journal ofbiomedical
semantics,vol.2,no.1,p.S2,2011.
[7]E.Beisswanger and U.Hahn,“Towards valid and reusable referencebasic quality checks for ontology alignments and their
application to three different reference data sets,”Journal ofbiomedical
semantics,vol.3,no.1,p.S4,2012.
[8]I.F.Cruz,C.Stroe,and M.Palmonari,“Interactive user feedback in
ontology matching using signature vectors,”in 2012IEEE 28th International
Conference on Data Engineering.IEEE,2012,pp.1321–1324.
[9]P.Lambrix andR.Kaliyaperumal,“A session-based approach for
aligning large ontologies,”in Extended Semantic Web Conference.Springer,2013,
pp.46–60.
[10]A.Groβ,J.C.Dos Reis,M.Hartung,C.Pruski,and E.Rahm,“Semi-automatic
adaptation of mappings between life science ontologies,”in International
Conference on Data Integration inthe Life Sciences.Springer,2013,pp.90–104.
[11]N.F.Noy,M.A.Musen et al.,“Algorithm and tool for automated
ontology merging and alignment,”in Proceedings of the 17th National
Conference on Artificial Intelligence(AAAI-00).Available as SMI technical
report SMI-2000-0831,2000.
The content of the invention
The technical problem to be solved in the present invention, it is to provide a kind of interactive Ontology Matching method based on evolution algorithm,
The time point of user's intervention can be adaptively determined, automatically determines a limited number of Ontology Matching Candidate Sets by subscriber checking,
And the result of subscriber checking is spread to realize the purpose for the value for maximizing subscriber checking result.
What the inventive method was realized in:A kind of interactive Ontology Matching method based on evolution algorithm, including:
Body divides the stage, large-scale body is divided into small-scale body piecemeal so that follow-up Ontology Matching
Process is carried out in body piecemeal;
In the Ontology Matching stage based on evolution algorithm, the Ontology Matching process of automation is realized using evolution algorithm, it is adaptive
Ground is answered to determine the time point of user's intervention;
Subscriber checking and assay diffusion phase, the candidate's Ontology Matching result automatically determined is presented to user's inspection
Test, remaining candidate's Ontology Matching result is automatically processed during subscriber checking, finally spread believable subscriber checking knot
Fruit;
Body divided-fit surface result integrates and evaluation phase, integrates different body divided-fit surface results, and utilize body
The f measurements of final result are calculated with reference to matching result.
Further, the detailed process in the body division stage is:
(1) dispersiveness and the disequilibrium for measuring body construction by SemanticAccuaracy first are reliable to select
The stronger body of property is as source body;
(2) source body and then using extension from the body partitioning algorithm of SCAN algorithms is divided into source body piecemeal;
(3) to each source body piecemeal, measured using conceptual dependency degree to determine its similar target body piecemeal.
Further, the Ontology Matching stage based on evolution algorithm includes modeling process and matching process;
The modeling process is specifically:
S11, body O is defined as to O={ C, P, I }, wherein C, P and I represent the concept set in body, property set respectively
Conjunction and example collection, wherein concept, attribute and example are referred to as the entity of body;Ontology Matching result A is an Entities Matching pair
Set, each Entities Matching to that can be expressed as a four-tuple { e, e', n, rel }, wherein e and e' represent respectively it is first this
The entity of body and target body, n are the confidence values of e and e' relations, and rel is the equivalence relation between e and e';
S12, given Ontology Matching result A, its quality are measured by f (A):
Wherein, | A | be the matching in A to quantity, MF (A) be calculate A MatchFmeasure values, δiIt is i-th in A
The Similarity value of matching pair, α ∈ [0,1] are the recall ratio and precision ratio main modulation parameter for weighing Ontology Matching result;
S13, given source body piecemeal OsrcWith target body piecemeal Otgt, the optimization mould of design single goal Ontology Matching problem
Type is as follows:
Wherein, fi(X), i=1,2 ..., m calculate f () value of i-th of Ontology Matching device result, | Osrc| and | Otgt| point
Biao Shi not body OsrcAnd OtgtEntity sets radix, xi, i=1,2 ..., | Osrc| represent i-th of Entities Matching pair;
The matching process is specifically:
S21, placement algorithm control parameter, including numerical precision, population scale, select probability, crossover probability and variation are general
Rate;
S22, random initializtion population, and initialization of the fitness value highest individual as elite solution is selected in population
Value;
S23, into evolutionary process, evaluate each individual fitness value f () in population first, selected by roulette wheel
Select, and intersected by single-point and generate population of future generation with selection after the operation of single-point variance criteria evolutionary operator;
S24, the fitness value f () for reappraising population simultaneously attempt to update elite solution, until meeting end condition, and export
Body divided-fit surface result, if continuously some generations can not update and evolve elite solution, user is now allowed to intervene entering for bootstrap algorithm
Change direction.
Further, the detailed process of the subscriber checking and assay diffusion phase is:
(1) first by Entities Matching in elite solution to being " problem " matching near 0.4 in similarity threshold to allowing user
Examine;
(2) and then for an entity in target body there is matching corresponding to multiple source body entities, allow user to select it
In one correctly matching pair, remaining matching pair confidence level be arranged to 0;
(3) result crossed finally for subscriber checking, matching pair credible of the Similarity value higher than threshold value 0.9 is diffused into
In its periphery concept.
Further, the α values are proportional with recall ratio, are inversely proportional with precision ratio.The suggestion value of the α is 0.35.
In addition, the present invention also provides a kind of computer equipment, including memory, processor and storage are on a memory and can
The computer program run on a processor, the mistake of the invention described above method is able to carry out during the computing device described program
Journey.
The invention has the advantages that:
(1) opportunity of user's intervention is adaptively determined using evolution algorithm, reduces unnecessary user mutual;
(2) automatically according to subscriber checking result treatment residue candidate matches pair, reduce the workload of user, improve use
The efficiency of family checkout procedure;
(3) Entities Matching higher than threshold value for crossing subscriber checking is to being diffused into the concept of periphery, to reduce the use of mistake
The harmful effect that family assay is brought, the effect of Ontology Matching process is improved while the value of subscriber checking result is maximized
The quality of rate and final Ontology Matching result.
Brief description of the drawings
The present invention is further illustrated in conjunction with the embodiments with reference to the accompanying drawings.
Fig. 1 is the inventive method execution flow chart.
Fig. 2 is the execution flow chart of subscriber checking and assay diffusion phase in the inventive method.
Embodiment
Refer to shown in Fig. 1, the interactive Ontology Matching method of the invention based on evolution algorithm, including body division rank
Section, the Ontology Matching stage based on evolution algorithm, subscriber checking and assay diffusion phase, and body divided-fit surface result
Integrated and evaluation phase.
The body divides the stage, large-scale body is divided into small-scale body piecemeal so that follow-up body
Matching process is carried out in body piecemeal;Its detailed process is:
(1) SemanticAccuaracy (S á nchez D, Batet M, Mart í nez S, et are passed through first
al.Semantic variance:an intuitive measure for ontology accuracy evaluation
[J].Engineering Applications ofArtificial Intelligence,2015,39:89-99) measure body
The dispersiveness of structure selects the stronger body of reliability as source body with disequilibrium;
(2) source body and then using extension from the body partitioning algorithm of SCAN algorithms is divided into source body piecemeal;
(3) to each source body piecemeal, measured using conceptual dependency degree to determine its similar target body piecemeal
(X.Xue,J.Pan,A segment-based approach for large-scale ontology matching,
Knowledge and Information Systems(2017)1–18)。
The Ontology Matching stage based on evolution algorithm, the Ontology Matching process of automation is realized using evolution algorithm,
Adaptively determine the time point of user's intervention;The Ontology Matching stage based on evolution algorithm includes modeling process and matching
Process;
The modeling process is specifically:
S11, body O is defined as to O={ C, P, I }, wherein C, P and I represent the concept set in body, property set respectively
Conjunction and example collection, wherein concept, attribute and example are referred to as the entity of body;Ontology Matching result A is an Entities Matching pair
Set, each Entities Matching to that can be expressed as a four-tuple { e, e', n, rel }, wherein e and e' represent respectively it is first this
The entity of body and target body, n are the confidence values of e and e' relations, and rel is the equivalence relation between e and e';
S12, given Ontology Matching result A, its quality are measured by f (A):
Wherein, | A | be the matching in A to quantity, MF (A) be calculate A MatchFmeasure values, δiIt is i-th in A
The Similarity value of matching pair, α ∈ [0,1] are the recall ratio and precision ratio main modulation parameter for weighing Ontology Matching result;Institute
It is proportional with recall ratio to state α values, is inversely proportional with precision ratio;The suggestion value of the α is 0.35;
S13, given source body piecemeal OsrcWith target body piecemeal Otgt, the optimization mould of design single goal Ontology Matching problem
Type is as follows:
Wherein fi(X), i=1,2 ..., m calculate f () value of i-th of Ontology Matching device result, | Osrc| and | Otgt| respectively
Represent body OsrcAnd OtgtEntity sets radix, xi, i=1,2 ..., | Osrc| represent i-th of Entities Matching pair;
The matching process is specifically:
S21, placement algorithm control parameter, including numerical precision, population scale, select probability, crossover probability and variation are general
Rate;
S22, random initializtion population, and initialization of the fitness value highest individual as elite solution is selected in population
Value;
S23, into evolutionary process, evaluate each individual fitness value f () in population first, selected by roulette wheel
Select, and intersected by single-point and generate population of future generation with selection after the operation of single-point variance criteria evolutionary operator;
S24, the fitness value f () for reappraising population simultaneously attempt to update elite solution, until meeting end condition, and export
Body divided-fit surface result, if continuously some generations can not update and evolve elite solution, user is now allowed to intervene entering for bootstrap algorithm
Change direction.
The subscriber checking and assay diffusion phase, the candidate's Ontology Matching result automatically determined is presented to user
Examine, remaining candidate's Ontology Matching result is automatically processed during subscriber checking, finally spreads believable subscriber checking
As a result;Its detailed process is:
(1) first by Entities Matching in elite solution to being " problem " matching near 0.4 in similarity threshold to allowing user
Examine;
(2) and then for an entity in target body there is matching corresponding to multiple source body entities, allow user to select it
In one correctly matching pair, remaining matching pair confidence level be arranged to 0;
(3) result crossed finally for subscriber checking, matching pair credible of the Similarity value higher than threshold value 0.9 is diffused into
In its periphery concept;
For example, e and e' are found out respectively in respective Ontological concept to { e, e', 0.92 ,=} for the matching that subscriber checking is crossed
All father's concept set Sup in architectureeAnd Supe'IfI=1,2 ..., | Supe| andJ=1,
2,...,|Supe'|, it has been the entity pair matched, then its confidence level has been added 0.3.
The body divided-fit surface result integrates and evaluation phase, and different body divided-fit surface knots are integrated by greedy algorithm
Fruit, and the f measurements of the reference matching result calculating final result using body.
It should be noted that:The body partitioning algorithm of the present invention can also be replaced by other body partitioning algorithms;
The measurement technology for the approximate Ontology Matching result that the present invention uses can also use other approximate Ontology Matching result measurement skills
Art replaces;The evolution algorithm that the present invention uses can also be replaced using other swarm intelligence algorithms.In addition, the present invention also provides one kind
Computer equipment, including memory, processor and storage are on a memory and the computer program that can run on a processor, institute
The process of the invention described above method is able to carry out when stating computing device described program.
Although the foregoing describing the embodiment of the present invention, those familiar with the art should manage
Solution, the specific embodiment described by us are merely exemplary, rather than for the restriction to the scope of the present invention, are familiar with this
The equivalent modification and change that the technical staff in field is made in the spirit according to the present invention, should all cover the present invention's
In scope of the claimed protection.
Claims (12)
- A kind of 1. interactive Ontology Matching method based on evolution algorithm, it is characterised in that:Including:Body divides the stage, large-scale body is divided into small-scale body piecemeal so that follow-up Ontology Matching process It is to be carried out in body piecemeal;In the Ontology Matching stage based on evolution algorithm, the Ontology Matching process of automation is realized using evolution algorithm, adaptively Determine the time point of user's intervention;Subscriber checking and assay diffusion phase, the candidate's Ontology Matching result automatically determined is presented to subscriber checking, Remaining candidate's Ontology Matching result is automatically processed during subscriber checking, finally spreads believable subscriber checking result;Body divided-fit surface result integrates and evaluation phase, integrates different body divided-fit surface results, and utilize the reference of body Matching result calculates the f measurements of final result.
- A kind of 2. interactive Ontology Matching method based on evolution algorithm according to claim 1, it is characterised in that:It is described Body division the stage detailed process be:(1) first by SemanticAccuaracy measure body construction dispersiveness and disequilibrium come select reliability compared with Strong body is as source body;(2) source body and then using extension from the body partitioning algorithm of SCAN algorithms is divided into source body piecemeal;(3) to each source body piecemeal, measured using conceptual dependency degree to determine its similar target body piecemeal.
- A kind of 3. interactive Ontology Matching method based on evolution algorithm according to claim 1, it is characterised in that:It is described The Ontology Matching stage based on evolution algorithm includes modeling process and matching process;The modeling process is specifically:S11, body O is defined as to O={ C, P, I }, wherein C, P and I represent the concept set in body respectively, attribute set and Example collection, wherein concept, attribute and example are referred to as the entity of body;Ontology Matching result A is the collection of an Entities Matching pair Close, each Entities Matching to that can be expressed as a four-tuple { e, e', n, rel }, wherein e and e' represent respectively meta-ontology and The entity of target body, n are the confidence values of e and e' relations, and rel is the equivalence relation between e and e';S12, given Ontology Matching result A, its quality are measured by f (A):<mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>A</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>&alpha;</mi> <mo>&times;</mo> <mi>M</mi> <mi>F</mi> <mrow> <mo>(</mo> <mi>A</mi> <mo>)</mo> </mrow> <mo>+</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&alpha;</mi> <mo>)</mo> </mrow> <mo>&times;</mo> <mfrac> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mo>|</mo> <mi>A</mi> <mo>|</mo> </mrow> </msubsup> <msub> <mi>&delta;</mi> <mi>i</mi> </msub> </mrow> <mrow> <mo>|</mo> <mi>A</mi> <mo>|</mo> </mrow> </mfrac> <mo>&Element;</mo> <mo>&lsqb;</mo> <mn>0</mn> <mo>,</mo> <mn>1</mn> <mo>&rsqb;</mo> <mo>,</mo> </mrow>Wherein, | A | be the matching in A to quantity, MF (A) be calculate A MatchFmeasure values, δiIt is i-th of matching pair in A Similarity value, α ∈ [0,1] are the recall ratio and precision ratio main modulation parameter for weighing Ontology Matching result;S13, given source body piecemeal OsrcWith target body piecemeal Otgt, design the Optimized model of single goal Ontology Matching problem such as Under:<mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mi>min</mi> </mtd> <mtd> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>X</mi> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>s</mi> <mo>.</mo> <mi>t</mi> <mo>.</mo> </mrow> </mtd> <mtd> <mrow> <mi>X</mi> <mo>=</mo> <msup> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>,</mo> <mo>...</mo> <mo>,</mo> <msub> <mi>x</mi> <mrow> <mo>|</mo> <mi>s</mi> <mi>r</mi> <mi>c</mi> <mo>|</mo> </mrow> </msub> <mo>)</mo> </mrow> <mi>T</mi> </msup> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow></mrow> </mtd> <mtd> <mrow> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>&Element;</mo> <mo>{</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mo>|</mo> <msub> <mi>O</mi> <mrow> <mi>t</mi> <mi>g</mi> <mi>t</mi> </mrow> </msub> <mo>|</mo> <mo>}</mo> </mrow> </mtd> </mtr> </mtable> </mfenced>Wherein, fi(X), i=1,2 ..., m calculate f () value of i-th of Ontology Matching device result, | Osrc| and | Otgt| table respectively Show body OsrcAnd OtgtEntity sets radix, xi, i=1,2 ..., | Osrc| represent i-th of Entities Matching pair;The matching process is specifically:S21, placement algorithm control parameter, including numerical precision, population scale, select probability, crossover probability and mutation probability;S22, random initializtion population, and initialization value of the fitness value highest individual as elite solution is selected in population;S23, into evolutionary process, evaluate each individual fitness value f () in population first, selected by roulette wheel, with And intersected by single-point and generate population of future generation with selection after the operation of single-point variance criteria evolutionary operator;S24, the fitness value f () for reappraising population simultaneously attempt to update elite solution, until meeting end condition, and export body Divided-fit surface result, if continuously some generations can not update and evolve elite solution, user is now allowed to intervene the evolution side of bootstrap algorithm To progress subscriber checking and assay diffusion phase.
- A kind of 4. interactive Ontology Matching method based on evolution algorithm according to claim 1 or 3, it is characterised in that: The detailed process of the subscriber checking and assay diffusion phase is:(1) first by Entities Matching in elite solution to being " problem " matching near 0.4 in similarity threshold to allowing subscriber checking;(2) and then for an entity in target body there is matching corresponding to multiple source body entities, allow user to select wherein one Individual correctly matching pair, the confidence level of remaining matching pair are arranged to 0;(3) result crossed finally for subscriber checking, matching pair credible of the Similarity value higher than threshold value 0.9 is diffused into its week In the concept of side.
- A kind of 5. interactive Ontology Matching method based on evolution algorithm according to claim 3, it is characterised in that:It is described α values are proportional with recall ratio, are inversely proportional with precision ratio.
- A kind of 6. interactive Ontology Matching method based on evolution algorithm according to claim 3 or 5, it is characterised in that: The value of the α is 0.35.
- 7. a kind of computer equipment, including memory, processor and storage are on a memory and the meter that can run on a processor Calculation machine program, it is characterised in that realize procedure below during the computing device described program:Body divides the stage, large-scale body is divided into small-scale body piecemeal so that follow-up Ontology Matching process It is to be carried out in body piecemeal;In the Ontology Matching stage based on evolution algorithm, the Ontology Matching process of automation is realized using evolution algorithm, adaptively Determine the time point of user's intervention;Subscriber checking and assay diffusion phase, the candidate's Ontology Matching result automatically determined is presented to subscriber checking, Remaining candidate's Ontology Matching result is automatically processed during subscriber checking, finally spreads believable subscriber checking result;Body divided-fit surface result integrates and evaluation phase, integrates different body divided-fit surface results, and utilize the reference of body Matching result calculates the f measurements of final result.
- A kind of 8. computer equipment according to claim 7, it is characterised in that:The detailed process in the body division stage It is:(1) first by SemanticAccuaracy measure body construction dispersiveness and disequilibrium come select reliability compared with Strong body is as source body;(2) source body and then using extension from the body partitioning algorithm of SCAN algorithms is divided into source body piecemeal;(3) to each source body piecemeal, measured using conceptual dependency degree to determine its similar target body piecemeal.
- A kind of 9. computer equipment according to claim 7, it is characterised in that:The Ontology Matching based on evolution algorithm Stage includes modeling process and matching process;The modeling process is specifically:S11, body O is defined as to O={ C, P, I }, wherein C, P and I represent the concept set in body respectively, attribute set and Example collection, wherein concept, attribute and example are referred to as the entity of body;Ontology Matching result A is the collection of an Entities Matching pair Close, each Entities Matching to that can be expressed as a four-tuple { e, e', n, rel }, wherein e and e' represent respectively meta-ontology and The entity of target body, n are the confidence values of e and e' relations, and rel is the equivalence relation between e and e';S12, given Ontology Matching result A, its quality are measured by f (A):<mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>A</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>&alpha;</mi> <mo>&times;</mo> <mi>M</mi> <mi>F</mi> <mrow> <mo>(</mo> <mi>A</mi> <mo>)</mo> </mrow> <mo>+</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&alpha;</mi> <mo>)</mo> </mrow> <mo>&times;</mo> <mfrac> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mo>|</mo> <mi>A</mi> <mo>|</mo> </mrow> </msubsup> <msub> <mi>&delta;</mi> <mi>i</mi> </msub> </mrow> <mrow> <mo>|</mo> <mi>A</mi> <mo>|</mo> </mrow> </mfrac> <mo>&Element;</mo> <mo>&lsqb;</mo> <mn>0</mn> <mo>,</mo> <mn>1</mn> <mo>&rsqb;</mo> <mo>,</mo> </mrow>Wherein, | A | be the matching in A to quantity, MF (A) be calculate A MatchFmeasure values, δiIt is i-th of matching pair in A Similarity value, α ∈ [0,1] are the recall ratio and precision ratio main modulation parameter for weighing Ontology Matching result;S13, given source body piecemeal OsrcWith target body piecemeal Otgt, design the Optimized model of single goal Ontology Matching problem such as Under:<mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mi>min</mi> </mtd> <mtd> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>X</mi> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>s</mi> <mo>.</mo> <mi>t</mi> <mo>.</mo> </mrow> </mtd> <mtd> <mrow> <mi>X</mi> <mo>=</mo> <msup> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>,</mo> <mo>...</mo> <mo>,</mo> <msub> <mi>x</mi> <mrow> <mo>|</mo> <mi>s</mi> <mi>r</mi> <mi>c</mi> <mo>|</mo> </mrow> </msub> <mo>)</mo> </mrow> <mi>T</mi> </msup> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow></mrow> </mtd> <mtd> <mrow> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>&Element;</mo> <mo>{</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mo>|</mo> <msub> <mi>O</mi> <mrow> <mi>t</mi> <mi>g</mi> <mi>t</mi> </mrow> </msub> <mo>|</mo> <mo>}</mo> </mrow> </mtd> </mtr> </mtable> </mfenced>Wherein, fi(X), i=1,2 ..., m calculate f () value of i-th of Ontology Matching device result, | Osrc| and | Otgt| table respectively Show body OsrcAnd OtgtEntity sets radix, xi, i=1,2 ..., | Osrc| represent i-th of Entities Matching pair;The matching process is specifically:S21, placement algorithm control parameter, including numerical precision, population scale, select probability, crossover probability and mutation probability;S22, random initializtion population, and initialization value of the fitness value highest individual as elite solution is selected in population;S23, into evolutionary process, evaluate each individual fitness value f () in population first, selected by roulette wheel, with And intersected by single-point and generate population of future generation with selection after the operation of single-point variance criteria evolutionary operator;S24, the fitness value f () for reappraising population simultaneously attempt to update elite solution, until meeting end condition, and export body Divided-fit surface result, if continuously some generations can not update and evolve elite solution, user is now allowed to intervene the evolution side of bootstrap algorithm To into the subscriber checking and assay diffusion phase.
- A kind of 10. computer equipment according to claim 7 or 9, it is characterised in that:The subscriber checking and assay The detailed process of diffusion phase is:(1) first by Entities Matching in elite solution to being " problem " matching near 0.4 in similarity threshold to allowing subscriber checking;(2) and then for an entity in target body there is matching corresponding to multiple source body entities, allow user to select wherein one Individual correctly matching pair, the confidence level of remaining matching pair are arranged to 0;(3) result crossed finally for subscriber checking, matching pair credible of the Similarity value higher than threshold value 0.9 is diffused into its week In the concept of side.
- A kind of 11. interactive Ontology Matching method based on evolution algorithm according to claim 9, it is characterised in that:Institute It is proportional with recall ratio to state α values, is inversely proportional with precision ratio.
- 12. a kind of interactive Ontology Matching method based on evolution algorithm according to claim 9 or 11, its feature exist In:The value of the α is 0.35.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711070247.0A CN107766556B (en) | 2017-11-03 | 2017-11-03 | Interactive ontology matching method based on evolutionary algorithm and computer equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711070247.0A CN107766556B (en) | 2017-11-03 | 2017-11-03 | Interactive ontology matching method based on evolutionary algorithm and computer equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107766556A true CN107766556A (en) | 2018-03-06 |
CN107766556B CN107766556B (en) | 2021-07-30 |
Family
ID=61273656
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711070247.0A Active CN107766556B (en) | 2017-11-03 | 2017-11-03 | Interactive ontology matching method based on evolutionary algorithm and computer equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107766556B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112860913A (en) * | 2021-02-24 | 2021-05-28 | 广州汇通国信科技有限公司 | Ontology creation method of knowledge graph |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103631882A (en) * | 2013-11-14 | 2014-03-12 | 北京邮电大学 | Semantization service generation system and method based on graph mining technique |
CN103810388A (en) * | 2014-02-19 | 2014-05-21 | 福建工程学院 | Large-scale ontology mapping method based on partitioning technology oriented towards mapping |
US20150149191A1 (en) * | 2013-11-27 | 2015-05-28 | University-Industry Cooperation Group Of Kyung Hee University | Apparatus and method for matching large-scale biomedical ontologies |
CN104866626A (en) * | 2015-06-15 | 2015-08-26 | ***通信集团黑龙江有限公司 | Method and device for recommending telecommunication service |
WO2016093405A1 (en) * | 2014-12-12 | 2016-06-16 | University-Industry Cooperation Group Of Kyung Hee University | Adaptive semantic based mediation system and method |
CN106227798A (en) * | 2016-07-21 | 2016-12-14 | 福州职业技术学院 | A kind of Ontology Matching method based on compact Cooperative Evolutionary Algorithm |
CN106855886A (en) * | 2016-12-26 | 2017-06-16 | 武汉大学 | The position concept structuring extracting method of the conspicuousness structure based on location expression |
CN107193858A (en) * | 2017-03-28 | 2017-09-22 | 福州金瑞迪软件技术有限公司 | Towards the intelligent Service application platform and method of multi-source heterogeneous data fusion |
-
2017
- 2017-11-03 CN CN201711070247.0A patent/CN107766556B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103631882A (en) * | 2013-11-14 | 2014-03-12 | 北京邮电大学 | Semantization service generation system and method based on graph mining technique |
US20150149191A1 (en) * | 2013-11-27 | 2015-05-28 | University-Industry Cooperation Group Of Kyung Hee University | Apparatus and method for matching large-scale biomedical ontologies |
CN103810388A (en) * | 2014-02-19 | 2014-05-21 | 福建工程学院 | Large-scale ontology mapping method based on partitioning technology oriented towards mapping |
WO2016093405A1 (en) * | 2014-12-12 | 2016-06-16 | University-Industry Cooperation Group Of Kyung Hee University | Adaptive semantic based mediation system and method |
CN104866626A (en) * | 2015-06-15 | 2015-08-26 | ***通信集团黑龙江有限公司 | Method and device for recommending telecommunication service |
CN106227798A (en) * | 2016-07-21 | 2016-12-14 | 福州职业技术学院 | A kind of Ontology Matching method based on compact Cooperative Evolutionary Algorithm |
CN106855886A (en) * | 2016-12-26 | 2017-06-16 | 武汉大学 | The position concept structuring extracting method of the conspicuousness structure based on location expression |
CN107193858A (en) * | 2017-03-28 | 2017-09-22 | 福州金瑞迪软件技术有限公司 | Towards the intelligent Service application platform and method of multi-source heterogeneous data fusion |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112860913A (en) * | 2021-02-24 | 2021-05-28 | 广州汇通国信科技有限公司 | Ontology creation method of knowledge graph |
CN112860913B (en) * | 2021-02-24 | 2024-03-08 | 广州汇通国信科技有限公司 | Ontology creation method of knowledge graph |
Also Published As
Publication number | Publication date |
---|---|
CN107766556B (en) | 2021-07-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ros et al. | A machine learning approach for semi-automated search and selection in literature studies | |
Xue et al. | Collaborative ontology matching based on compact interactive evolutionary algorithm | |
CN106663038B (en) | Feature processing recipe for machine learning | |
CN106575246B (en) | Machine learning service | |
Wang et al. | Compositional vector space models for improved bug localization | |
Odenbaugh et al. | Buyer beware: Robustness analyses in economics and biology | |
Marcus et al. | Bao: Learning to steer query optimizers | |
CN110462612A (en) | The method and apparatus for carrying out machine learning using the network at network node with ageng and ranking then being carried out to network node | |
Jafari et al. | Lithium-ion battery estimation in online framework using extreme gradient boosting machine learning approach | |
Simhadri et al. | Results of the NeurIPS’21 challenge on billion-scale approximate nearest neighbor search | |
CN109117380A (en) | A kind of method for evaluating software quality, device, equipment and readable storage medium storing program for executing | |
Wen et al. | Robust least squares support vector machine based on recursive outlier elimination | |
Csar et al. | Winner determination in huge elections with mapreduce | |
Berggren | SGV 3.0-a fast detector simulation | |
Bendík et al. | Approximate counting of minimal unsatisfiable subsets | |
Shen et al. | Multi-scale parameter identification of lithium-ion battery electric models using a PSO-LM algorithm | |
Sun et al. | SMT solver validation empowered by large pre-trained language models | |
Liu et al. | A data-driven approach for online inter-area oscillatory stability assessment of power systems based on random bits forest considering feature redundancy | |
Marques et al. | An overview of methods and technologies for estimating battery state of charge in electric vehicles | |
Xu et al. | Aging characteristics and state-of-health estimation of retired batteries: an electrochemical impedance spectroscopy perspective | |
CN107766556A (en) | A kind of interactive Ontology Matching method and computer equipment based on evolution algorithm | |
Gong et al. | A novel method of grouping target paths for parallel programs | |
Mandal et al. | Regret theory based three-way conflict analysis model under q-rung orthopair fuzzy information: studies with parameter and three-way decision-making-based approaches | |
Zheng et al. | Comparing alternative biometric models with and without gene-by-measured environment interaction in behavior genetic designs: Statistical operating characteristics | |
Matczak et al. | Intelligent index tuning using reinforcement learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |