A hybrid approach to NER by integrating manual rules into MEMM

Moshe Fresko; Binyamin Rozenfeld; Ronen Feldman

A hybrid approach to NER by integrating manual rules into MEMM

Moshe Fresko, Binyamin Rozenfeld, Ronen Feldman

Department of Computer Science

Bar-Ilan University

Research output: Contribution to conference › Paper › peer-review

2 Scopus citations

Abstract

This paper describes a framework for defining domain specific Feature Functions in a user friendly form to be used in a Maximum Entropy Markov Model (MEMM) for the Named Entity Recognition (NER) task. Our system called MERGE allows defining general Feature Function Templates, as well as Linguistic Rules incorporated into the classifier. The simple way of translating these rules into specific feature functions are shown. We show that MERGE can perform better from both purely machine learning based systems and purely-knowledge based approaches by some small expert interaction of rule-tuning.

Original language	English
State	Published - 2006
Event	9th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2006 - Fort Lauderdale, FL, United States Duration: 4 Jan 2006 → 6 Jan 2006

Conference

Conference	9th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2006
Country/Territory	United States
City	Fort Lauderdale, FL
Period	4/01/06 → 6/01/06

Keywords

Information extraction
Machine learning
Named entity recognition

Cite this

@conference{e760c8c8aa3b4f2799187acebafd2996,

title = "A hybrid approach to NER by integrating manual rules into MEMM",

abstract = "This paper describes a framework for defining domain specific Feature Functions in a user friendly form to be used in a Maximum Entropy Markov Model (MEMM) for the Named Entity Recognition (NER) task. Our system called MERGE allows defining general Feature Function Templates, as well as Linguistic Rules incorporated into the classifier. The simple way of translating these rules into specific feature functions are shown. We show that MERGE can perform better from both purely machine learning based systems and purely-knowledge based approaches by some small expert interaction of rule-tuning.",

keywords = "Information extraction, Machine learning, Named entity recognition",

author = "Moshe Fresko and Binyamin Rozenfeld and Ronen Feldman",

year = "2006",

language = "אנגלית",

note = "9th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2006 ; Conference date: 04-01-2006 Through 06-01-2006",

}

TY - CONF

T1 - A hybrid approach to NER by integrating manual rules into MEMM

AU - Fresko, Moshe

AU - Rozenfeld, Binyamin

AU - Feldman, Ronen

PY - 2006

Y1 - 2006

N2 - This paper describes a framework for defining domain specific Feature Functions in a user friendly form to be used in a Maximum Entropy Markov Model (MEMM) for the Named Entity Recognition (NER) task. Our system called MERGE allows defining general Feature Function Templates, as well as Linguistic Rules incorporated into the classifier. The simple way of translating these rules into specific feature functions are shown. We show that MERGE can perform better from both purely machine learning based systems and purely-knowledge based approaches by some small expert interaction of rule-tuning.

AB - This paper describes a framework for defining domain specific Feature Functions in a user friendly form to be used in a Maximum Entropy Markov Model (MEMM) for the Named Entity Recognition (NER) task. Our system called MERGE allows defining general Feature Function Templates, as well as Linguistic Rules incorporated into the classifier. The simple way of translating these rules into specific feature functions are shown. We show that MERGE can perform better from both purely machine learning based systems and purely-knowledge based approaches by some small expert interaction of rule-tuning.

KW - Information extraction

KW - Machine learning

KW - Named entity recognition

UR - http://www.scopus.com/inward/record.url?scp=84864536186&partnerID=8YFLogxK

M3 - ???researchoutput.researchoutputtypes.contributiontoconference.paper???

AN - SCOPUS:84864536186

T2 - 9th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2006

Y2 - 4 January 2006 through 6 January 2006

ER -

A hybrid approach to NER by integrating manual rules into MEMM

Abstract

Conference

Keywords

Other files and links

Fingerprint

Cite this