Reviewing and evaluating Automatic Term Recognition techniques

Ioannis Korkontzelos; Loannis P. Klapaftis; Suresh Manandhar

Reviewing and evaluating Automatic Term Recognition techniques

Ioannis Korkontzelos, Loannis P. Klapaftis, Suresh Manandhar

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Automatic Term Recognition (ATR) is defined as the task of identifying domain specific terms from technical corpora. Termhood-based approaches measure the degree that a candidate term refers to a domain specific concept. Unithood-based approaches measure the attachment strength of a candidate term constituents. These methods have been evaluated using different, often incompatible evaluation schemes and datasets. This paper provides an overview and a thorough evaluation of state-of-the-art ATR methods, under a common evaluation framework, i.e. corpora and evaluation method. Our contributions are two-fold: (1) We compare a number of different ATR methods, showing that termhood-based methods achieve in general superior performance. (2) We show that the number of independent occurrences of a candidate term is the most effective source for estimating term nestedness, improving ATR performance.

Original language	English
Title of host publication	ADVANCES IN NATURAL LANGUAGE PROCESSING, PROCEEDINGS
Editors	B Nordstrom, A Ranta
Place of Publication	BERLIN
Publisher	Springer
Pages	248-259
Number of pages	12
Volume	5221 LNAI
ISBN (Print)	978-3-540-85286-5
Publication status	Published - 2008
Event	6th International Conference on Natural Language Processing - Gothenburg Duration: 25 Aug 2008 → 27 Aug 2008

Conference

Conference	6th International Conference on Natural Language Processing
City	Gothenburg
Period	25/08/08 → 27/08/08

Keywords

automatic term recognition
ATR
term extraction

Cite this

@inproceedings{38443be61af44253ab12d36f942c79b7,

title = "Reviewing and evaluating Automatic Term Recognition techniques",

abstract = "Automatic Term Recognition (ATR) is defined as the task of identifying domain specific terms from technical corpora. Termhood-based approaches measure the degree that a candidate term refers to a domain specific concept. Unithood-based approaches measure the attachment strength of a candidate term constituents. These methods have been evaluated using different, often incompatible evaluation schemes and datasets. This paper provides an overview and a thorough evaluation of state-of-the-art ATR methods, under a common evaluation framework, i.e. corpora and evaluation method. Our contributions are two-fold: (1) We compare a number of different ATR methods, showing that termhood-based methods achieve in general superior performance. (2) We show that the number of independent occurrences of a candidate term is the most effective source for estimating term nestedness, improving ATR performance.",

keywords = "automatic term recognition, ATR, term extraction",

author = "Ioannis Korkontzelos and Klapaftis, {Loannis P.} and Suresh Manandhar",

year = "2008",

language = "English",

isbn = "978-3-540-85286-5",

volume = "5221 LNAI",

pages = "248--259",

editor = "B Nordstrom and A Ranta",

booktitle = "ADVANCES IN NATURAL LANGUAGE PROCESSING, PROCEEDINGS",

publisher = "Springer",

address = "Germany",

note = "6th International Conference on Natural Language Processing ; Conference date: 25-08-2008 Through 27-08-2008",

}

TY - GEN

T1 - Reviewing and evaluating Automatic Term Recognition techniques

AU - Korkontzelos, Ioannis

AU - Klapaftis, Loannis P.

AU - Manandhar, Suresh

PY - 2008

Y1 - 2008

N2 - Automatic Term Recognition (ATR) is defined as the task of identifying domain specific terms from technical corpora. Termhood-based approaches measure the degree that a candidate term refers to a domain specific concept. Unithood-based approaches measure the attachment strength of a candidate term constituents. These methods have been evaluated using different, often incompatible evaluation schemes and datasets. This paper provides an overview and a thorough evaluation of state-of-the-art ATR methods, under a common evaluation framework, i.e. corpora and evaluation method. Our contributions are two-fold: (1) We compare a number of different ATR methods, showing that termhood-based methods achieve in general superior performance. (2) We show that the number of independent occurrences of a candidate term is the most effective source for estimating term nestedness, improving ATR performance.

AB - Automatic Term Recognition (ATR) is defined as the task of identifying domain specific terms from technical corpora. Termhood-based approaches measure the degree that a candidate term refers to a domain specific concept. Unithood-based approaches measure the attachment strength of a candidate term constituents. These methods have been evaluated using different, often incompatible evaluation schemes and datasets. This paper provides an overview and a thorough evaluation of state-of-the-art ATR methods, under a common evaluation framework, i.e. corpora and evaluation method. Our contributions are two-fold: (1) We compare a number of different ATR methods, showing that termhood-based methods achieve in general superior performance. (2) We show that the number of independent occurrences of a candidate term is the most effective source for estimating term nestedness, improving ATR performance.

KW - automatic term recognition

KW - ATR

KW - term extraction

UR - http://www.scopus.com/inward/record.url?scp=52149106920&partnerID=8YFLogxK

M3 - Conference contribution

SN - 978-3-540-85286-5

VL - 5221 LNAI

SP - 248

EP - 259

BT - ADVANCES IN NATURAL LANGUAGE PROCESSING, PROCEEDINGS

A2 - Nordstrom, B

A2 - Ranta, A

PB - Springer

CY - BERLIN

T2 - 6th International Conference on Natural Language Processing

Y2 - 25 August 2008 through 27 August 2008

ER -

Reviewing and evaluating Automatic Term Recognition techniques

Abstract

Conference

Keywords

Other files and links

Cite this