• Aucun résultat trouvé

Комбинированный подход к кросс-языковой идентификации сущностей (A Combined Approach to Cross-Language Identity Resolution)

N/A
N/A
Protected

Academic year: 2022

Partager "Комбинированный подход к кросс-языковой идентификации сущностей (A Combined Approach to Cross-Language Identity Resolution)"

Copied!
5
0
0

Texte intégral

(1)

-

© .. © ..

. .. !""

,

#

apanovich@iis.nsk.su mag@iis.nsk.su

" $ %$

& '", (' ) *+

+$$( $( + LOD +$$( $( +

$( +". - + +/$( ( *+

$( $( $(, $ / *& )&, '&'& )& . $( $( *+ * $

%" SpringerLink $

$ ( - !#.

1

- + , ' $(

!"" ( - !#) -$ ( -

!# [8] $ -$$( +$(

2$( (Linked Open Data, LOD) [10]. 2 ) +$( 2$( / * +$

$ (1), $ " "

(2), '" (3)

*) $( (4). [1]

$ $ / ' +$(

++), / $ $, +&' '"

RDF-$( RKBExplorer.com. 9$

+, +$$(

+$$( + +", " *"

*(

*$( , , )

*( +$( $ & "

. - + +/$( (

& %" $

*+ $( $(

$(, $ / *&

)&, '&'& )&

. " $

%$ ) '"

%" SpringerLink (http://link.springer.com/) -$ ( -

!#.

2

/$ % " +$ +"

' " % '", *, )

" owl:sameAs.

( * * ,

$ -$ ( - !#

)& % $, +& + ( ( . / , -$ ( *+&

+$$ , * , $ $ , *+&

+$$ ( / $( . <, / +* , $ *+ * +$$

$(, $ " %"

elibrary.ru? = ) & )&

" , , /&, + %" %"

+* , ( $ -$$ ( - !#.

% elibrary.ru / $* * +"

) ( )" + 10-15 , +" + )" ( , " . #

* /* elibrary.ru ) ( &", .. , 9.. @(, .. < ( ( ", +/ ( $ " ""

.

= SpringerLink $ $ % * . - $(, )+ $( ,

" , *

XVII !

DAMDID/RCDL’2015 «

" », #, 13-16 2015

(2)

/& ( - !#. - $(, / $ $ PDF ( )". / $ $ )" $, SpringerLink /

& - & )&

+, $ ( + *), ) " . *(, %"

+ , *+$ WorldCat.org [4], – +& * +$

, A&' *$(

, ( 9 < G, *( %$( . WorldCat.org ( -$$( +$(

2$(. 2$ WorldCat.org $ RDF- , + "

*+ schema.org + BiblioGraph (http://BiblioGraph.net).

=$ WordlCat.org +, % , / $ , ) +$$(

. #, '" WorldCat Identities, (http://www.worldcat.org/wcidentities/lccn- n80162678), /' $&'( ", $ " +*, ' &

. - / )& (

*( ..

) .. ( ) + # . % * , 1989 2012 , $ $ «$ + .. » ( ..

1988 ).

$$( , ) '" &

+$ $ « $ "$», + $ & %. ' ) $( " +$(

" OCLC

" < G (viaf.org), , , ++$ ) ( ( + (, +)(, ( + (. # , ) *+& * $

$, +*$ *$. @, , , "

, ) http://viaf.org/viaf/5347110, ,

$ $(, , $(

" * $ , $ $'$ )", ,

$ . "

$, * $

(http://viaf.org/viaf/196995053), /

$ $ )"

, $ % («! ( …»), $ $ $ / .

@ *, +& ), " " /

$ & $ *( +$(

, " $ $

& +$ .

+ " " ) '" +"

-$$( +$( $(

SILK[6]. = , *+$ VIAF DBLP,

$ ( [5, 7]. / &

/ $ &

+$$( ( +". [12, 13].

Q' " + ) '"

$(, + +$(

+" . ( +$$(

( / +$ $( , $ )". Q %" $ / * VIAF, ( %$ +$ &, , - T, ' * +/$ $ ). " $, + ) +$( , +$( $ (/

/ +, + (* +) ). % $ ) ( % ", ) ( , ('(

$( $(, +$

. # $ ) & $ )", + $( +/, ) $ +$

$. Q *)$ &

+$( ( ( / +*

' . ' ' &

$ $ ) , &&' + ), , , -+

[9, 14].

+$$( +$$( %

$ / $ (', , +$ $ / , , +$

+" " . % $

*+ ) '"

$" (, &'"

)"

)". ( % $ ' %"

SpringerLink. ! * * ),

$( + +$ * ( * , +$ « )"» +" $).

-' ( $ ) '" &'" :

(3)

x +$ (, ) $ % ( -

!# ( , " ),

& +/$ $ +$

'

" %"

SpringerLink.

x * ( - !# *+&

+$$ + +)",

%" SpringerLink + & +$$ + , ' + +) ' Google (Google translator.com). X /" ""

* SpringerLink + $ + ' $, +$ "

$ ( - !#.

' Jaro- Winkler [3]. *, ) + " +)"

/& , , + "

" / +) , &

+$ $ '". $(

*( $ +$ ', +$ (, -

!#).

x 2 ) * $ +"

+), + $ + ( -

!#.

x "$ * + &

$ ) $ $. @$

", $( $ +, & *, +'$ . '" ( +/: ' tf-idf [16] "

+, / LDA (Latent Dirichlet Allocation) [2].

LDA, / $ ' ) <* — Z".

$ ( , (

& ' ) [15].

x 2 ( ", $( + (/ + / ' &'(

, + +

NewgroupN, N-% $"

* + " $.

x 2 /" $ " + +, + + + $(

$ + . (

/ *, /&

, ++.

<) , & $, $ ), / ) + , " (W) ( / . ( / $ , / % +. + + . #, ) + 30 0.05.

$" +/ '

$ , (/

$ & / . / + $& &' . / = Temperature * SpringForce(d) * W SpringForceK ;

= Temperature * ElectricForce(d) / W * ElectricForceK;

SpringForce(D) = 2 * log(D);

ElectricForce(D) = 1 / d2, -% , W– ( / .

*+ * +/* *

* (" ), + &'" " $, * ) , " $, ' $.

3 $ " %

) '"

* ( .. , )" % ( .. , )(

- !#, &'( $" .

!+*$ $ $ %" $ * %"

" elibrary.ru.

Q $( SpringerLink, - $( * +*$" + A " ) )( (

$ +) *( )), ' * ).

@/ +*$ $ "

$ + 100 ( 3000 )") +, 80% , )( $ )

$, * * )$. $ $ +$ 70% .

# !. 1 + $ $, ''" + $ “.. ” ) %"

SpringerLink.com. $ "

( - !# $ +$

+$$ $ % ,

(4)

+$ (" . "

+$ +$$ $ $ +" $. ), + , +/&'" ), $ &

. $ ) & ), $(

$ +. $ / ) & ), $(

$ +. + “.. ” SpringerLink.com $ " 91 ). +

( 5 " / Andrei P. Ershov, * – Andrei Ershov, 84 - A.P. Ershov,

* – A.P. Yershov. # +(

$ + $ ( "$( ". + %( 91 )" . . * / 21 ), *$ * / ' * +$ ..

$.

!. 1. ), ) $ /' .. . *,

$( $ +, +$ $ ) .

" * ) 19 )" , 66 )"

*$( .. $(. +*

$( % * * * ) * *& * )" . . , /*

SpringerLink * )"

. . , /$(

% ( .. .

# " " $, ( ' " $ $(, "

$, )" + (& *, +&

93%.

4 &'(

" $ $ + (, *" $ ) $ $ )" ' + %" ).

* $ $ . =$ +,

$ $, * /* , &

$, % ) /* ( - !#. %( (, , $" + +$ ( )" + ( +$( , ) %(

( A*. ' %$ $ ('(

$( %) , + &'( A*

)& +*$

. @/ %$

$& + +

$( ' , / ), *+ +$(

" ) -) .

)

! $ " /

!TT ( _ 14-07-00386).

(5)

*

[1] Apanovich Z.V., Marchuk A.G. Experiments on using the LOD cloud datasets to enrich the content of a scientific knowledge base, P.Klinov and D.Mouromtsev (Eds.) KESW 2013, CCIS 394, Springer Verlag Berlin Heidelberg 2013, pp. 1-14.

[2] Blei D. M., Ng A., Jordan M. Latent Dirichlet allocationJournal of Machine Learning Research (3) 2003 pp. 993-1022.

[3] Cohen W. W., Ravikumar P. D., Fienberg S. E.: A Comparison of String Distance Metrics for Name- Matching Tasks. IIWeb 2003, pp. 73-78.

[4] Godby C. J., Denenberg R. Common Ground:

Exploring Compatibilities Between the Linked Data Models of the Library of Congress and OCLC

http://www.oclc.org/research/publications/2015/oc lcresearch-loc-linked-data-2015.html.

[5] Hickey, T. B., Toves J. A.. 2014. "Managing Ambiguity In VIAF" D-Lib Magazine 20 (July/August). doi:10.1045/july2014-

hickey.http://www.dlib.org/dlib/july14/hickey/07h ickey.html.

[6] Isele R., Jentzsch A., Bizer Ch. Silk Server - Adding missing Links while consuming Linked Data// 1st International Workshop on Consuming Linked Data (COLD 2010), Shanghai, November 2010.

[7] Ley M.: DBLP - Some Lessons Learned. PVLDB 2(2), 2009, pp. 1493-1500.

[8] Marchuk A.G., Marchuk P.A. Specific features of digital libraries construction with linked content.

Proc. of the RCDL’2010 Conf.– 2010. – P. 19–23.

(In Russian).

[9] Rogov A.A., Sidorov Yu. Vl. Statistical and Information-calculating Support of the Authorship Attribution of the Literary Works. Computer Data Analysis and Modeling: Robustness and Computer Intensive Methods: Proc. of the Sixth International Conference(September 10-14, 2001, Minsk).

Vol.2: K-S/ Edited by Prof. Dr. S. Aivazian, Prof.

Dr. Yu. Kharin and Prof. Dr. H. Rieder. Minsk:

BSU, 2001. – P. 187-192.

[10] Schultz A. et al. How to integrate LINKED DATA into your application //Semantic technology &

Business Conference, San Francisco, June 5, 2012.

http://mes-semantics.com/wp-

content/uploads/2012/09/Becker-etal-LDIF SemTechSanFrancisco.pdf.

[11] Steyvers M., Griffiths T. Probabilistic Topic Models Handbook of Latent Semantic Analysis.

2007.

[12] 9( . 9., #( . ., T . . - + $ ( +) $( / // . #X. .:

. (. – 2008. – @. 6, $. 1. – . 3–9.

[13] <+ . . +$

/ . . <+ , .

`.@ ", -. . < //

=$ : $

$ (, %$ )

(RCDL’2012) : . XIV. . .,

*-", 15–18 . 2012 . – *-" : +- «X », 2012. – . 360–369.

[14] jq 2.. Z +: !+

*+ )" ..

. X, .9: , N2, 2000, .115-126.

[15] http://snowball.tartarus.org/

[16] http://www.codeproject.com/Articles/12098/Term- frequency-Inverse-document-frequency-implemen

A Combined Approach to Cross-Language Identity Resolution

Zinaida V. Apanovich, Alexander G. Marchuk This paper describes experiments on the cross – language identity resolution problem that arises when the English-language LOD datasets are used to populate the content of a Russian scholarly knowledge base. One possible approach is the combined use of structured and text data, containing additional information and facilitating the identity resolution. The dataset of the Open Archive of the Russian Academy of Sciences and SpringerLink e-library are used as test examples.

Références

Documents relatifs

In monolingual part of our experiments we evaluated N-grams and a light stemmer on the Persian language and in cross language part we evaluated query translation and

In multilingual task (4 languages), the queries written in English were translating to French, Spanish and German. Once, we obtained the query translation, four different

A query object is a tuple hEAT, BoO, Keys, Li consisting of the expected answer type EAT, the BoO representation of the question, the set of relevant keywords used as query for

$SUHOLPLQDU\GLVWLQFWLRQFDQEH PDGHEHWZHHQWKHOH[LFDODQG QRQOH[LFDO SURVRGLF FKDUDFWHULVWLFV RI D ODQJXDJH /H[LFDO SDUDPHWHUV LQFOXGHGLVWLQFWLRQVRIVWUHVV RU DFFHQWWRQH DQGOHQJWK )UHQFK

In this paper, we propose a graph-based approach to multi-document summarization that integrates machine translation quality scores in the sentence extraction process.. We evaluate

Despite the substan- tive focus of this article on sport/terrorism, we argue that the broad epistemological, methodological and normative aspects of this critical social theory may

Using English as the main language for their services the examined community has been working on the creation of an inclusive atmosphere in which the linguistic translation