Efficient entity translation mining: A parallelized graph alignment approach

Gae Won You, Seung Won Hwang, Young In Song, Long Jiang, Zaiqing Nie

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

This article studies the problem of mining entity translation, specifically, mining English and Chinese name pairs. Existing efforts can be categorized into (a) transliteration-based approaches that leverage phonetic similarity and (b) corpus-based approaches that exploit bilingual cooccurrences. These approaches suffer from inaccuracy and scarcity, respectively. In clear contrast, we use under-leveraged resources of monolingual entity cooccurrences crawled from entity search engines, which are represented as two entityrelationship graphs extracted from two language corpora, respectively. Our problem is then abstracted as finding correct mappings across two graphs. To achieve this goal, we propose a holistic approach to exploiting both transliteration similarity and monolingual cooccurrences. This approach, which builds upon monolingual corpora, complements existing corpus-based work requiring scarce resources of parallel or comparable corpus while significantly boosting the accuracy of transliteration-based work. In addition, by parallelizing the mapping process on multicore architectures, we speed up the computation by more than 10 times per unit accuracy. We validated the effectiveness and efficiency of our proposed approach using real-life datasets.

Original languageEnglish
Article number25
JournalACM Transactions on Information Systems
Volume30
Issue number4
DOIs
Publication statusPublished - 2012 Nov 1

Fingerprint

Speech analysis
Search engines
Alignment
Graph
Resources

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Business, Management and Accounting(all)
  • Computer Science Applications

Cite this

You, Gae Won ; Hwang, Seung Won ; Song, Young In ; Jiang, Long ; Nie, Zaiqing. / Efficient entity translation mining : A parallelized graph alignment approach. In: ACM Transactions on Information Systems. 2012 ; Vol. 30, No. 4.
@article{c1756488dfe149c8815112d2f571e820,
title = "Efficient entity translation mining: A parallelized graph alignment approach",
abstract = "This article studies the problem of mining entity translation, specifically, mining English and Chinese name pairs. Existing efforts can be categorized into (a) transliteration-based approaches that leverage phonetic similarity and (b) corpus-based approaches that exploit bilingual cooccurrences. These approaches suffer from inaccuracy and scarcity, respectively. In clear contrast, we use under-leveraged resources of monolingual entity cooccurrences crawled from entity search engines, which are represented as two entityrelationship graphs extracted from two language corpora, respectively. Our problem is then abstracted as finding correct mappings across two graphs. To achieve this goal, we propose a holistic approach to exploiting both transliteration similarity and monolingual cooccurrences. This approach, which builds upon monolingual corpora, complements existing corpus-based work requiring scarce resources of parallel or comparable corpus while significantly boosting the accuracy of transliteration-based work. In addition, by parallelizing the mapping process on multicore architectures, we speed up the computation by more than 10 times per unit accuracy. We validated the effectiveness and efficiency of our proposed approach using real-life datasets.",
author = "You, {Gae Won} and Hwang, {Seung Won} and Song, {Young In} and Long Jiang and Zaiqing Nie",
year = "2012",
month = "11",
day = "1",
doi = "10.1145/2382438.2382444",
language = "English",
volume = "30",
journal = "ACM Transactions on Information Systems",
issn = "1046-8188",
publisher = "Association for Computing Machinery (ACM)",
number = "4",

}

Efficient entity translation mining : A parallelized graph alignment approach. / You, Gae Won; Hwang, Seung Won; Song, Young In; Jiang, Long; Nie, Zaiqing.

In: ACM Transactions on Information Systems, Vol. 30, No. 4, 25, 01.11.2012.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Efficient entity translation mining

T2 - A parallelized graph alignment approach

AU - You, Gae Won

AU - Hwang, Seung Won

AU - Song, Young In

AU - Jiang, Long

AU - Nie, Zaiqing

PY - 2012/11/1

Y1 - 2012/11/1

N2 - This article studies the problem of mining entity translation, specifically, mining English and Chinese name pairs. Existing efforts can be categorized into (a) transliteration-based approaches that leverage phonetic similarity and (b) corpus-based approaches that exploit bilingual cooccurrences. These approaches suffer from inaccuracy and scarcity, respectively. In clear contrast, we use under-leveraged resources of monolingual entity cooccurrences crawled from entity search engines, which are represented as two entityrelationship graphs extracted from two language corpora, respectively. Our problem is then abstracted as finding correct mappings across two graphs. To achieve this goal, we propose a holistic approach to exploiting both transliteration similarity and monolingual cooccurrences. This approach, which builds upon monolingual corpora, complements existing corpus-based work requiring scarce resources of parallel or comparable corpus while significantly boosting the accuracy of transliteration-based work. In addition, by parallelizing the mapping process on multicore architectures, we speed up the computation by more than 10 times per unit accuracy. We validated the effectiveness and efficiency of our proposed approach using real-life datasets.

AB - This article studies the problem of mining entity translation, specifically, mining English and Chinese name pairs. Existing efforts can be categorized into (a) transliteration-based approaches that leverage phonetic similarity and (b) corpus-based approaches that exploit bilingual cooccurrences. These approaches suffer from inaccuracy and scarcity, respectively. In clear contrast, we use under-leveraged resources of monolingual entity cooccurrences crawled from entity search engines, which are represented as two entityrelationship graphs extracted from two language corpora, respectively. Our problem is then abstracted as finding correct mappings across two graphs. To achieve this goal, we propose a holistic approach to exploiting both transliteration similarity and monolingual cooccurrences. This approach, which builds upon monolingual corpora, complements existing corpus-based work requiring scarce resources of parallel or comparable corpus while significantly boosting the accuracy of transliteration-based work. In addition, by parallelizing the mapping process on multicore architectures, we speed up the computation by more than 10 times per unit accuracy. We validated the effectiveness and efficiency of our proposed approach using real-life datasets.

UR - http://www.scopus.com/inward/record.url?scp=84871250455&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871250455&partnerID=8YFLogxK

U2 - 10.1145/2382438.2382444

DO - 10.1145/2382438.2382444

M3 - Article

AN - SCOPUS:84871250455

VL - 30

JO - ACM Transactions on Information Systems

JF - ACM Transactions on Information Systems

SN - 1046-8188

IS - 4

M1 - 25

ER -