東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Automated methods of auditing and us...

Fan, Jung-Wei.

Linked to FindBook

Google Book

Amazon

博客來

Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing./
Author:	Fan, Jung-Wei.
Description:	120 p.
Notes:	Source: Dissertation Abstracts International, Volume: 71-03, Section: B, page: 1426.
Contained By:	Dissertation Abstracts International71-03B.
Subject:	Biology, Bioinformatics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3400568
ISBN:	9781109673333

Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing.
Fan, Jung-Wei.

Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing. - 120 p.

Source: Dissertation Abstracts International, Volume: 71-03, Section: B, page: 1426.

Thesis (Ph.D.)--Columbia University, 2009.

Due to our cognitive nature of communicating in natural language, narrative information plays a critical role in storing and disseminating knowledge. In a knowledge-intensive domain such as biomedicine, the overhead to digest huge amount of texts in clinical reports, research literature, and consumer websites, is extremely demanding. Biomedical natural language processing (BioNLP) is an informatics specialty that aims to automatically analyze and restructure biomedical text into more digestible size and format so that it can be easily post-processed by humans or other automated programs. In order to handle the comprehensive lexical and semantic knowledge in biomedicine, BioNLP systems need to incorporate domain-specific terminology/ontology knowledge bases. In addition, using standardized lexical/semantic entities will benefit the interoperability between BioNLP systems and associated applications. However, two major issues have been observed as hindering the optimal use of terminology/ontology for BioNLP: First, the existing terminology/ontology knowledge bases are not customized for NLP purposes and contain problematic contents; Second, automated solutions for improving and using the knowledge bases are still inadequate and therefore limiting their use in BioNLP.

ISBN: 9781109673333Subjects--Topical Terms:

1018415
Biology, Bioinformatics.

Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing.
LDR:03396nam 2200289 4500 001 1401590
005 20111017084359.5
008 130515s2009 ||||||||||||||||| ||eng d
020 $a 9781109673333
035 $a (UMI)AAI3400568
035 $a AAI3400568
040 $a UMI $c UMI
100 1 $a Fan, Jung-Wei. $3 1680736
245 1 0 $a Automated methods of auditing and using terminology/ontology knowledge bases for natural language processing.
300 $a 120 p.
500 $a Source: Dissertation Abstracts International, Volume: 71-03, Section: B, page: 1426.
500 $a Adviser: Carol Friedman.
502 $a Thesis (Ph.D.)--Columbia University, 2009.
520 $a Due to our cognitive nature of communicating in natural language, narrative information plays a critical role in storing and disseminating knowledge. In a knowledge-intensive domain such as biomedicine, the overhead to digest huge amount of texts in clinical reports, research literature, and consumer websites, is extremely demanding. Biomedical natural language processing (BioNLP) is an informatics specialty that aims to automatically analyze and restructure biomedical text into more digestible size and format so that it can be easily post-processed by humans or other automated programs. In order to handle the comprehensive lexical and semantic knowledge in biomedicine, BioNLP systems need to incorporate domain-specific terminology/ontology knowledge bases. In addition, using standardized lexical/semantic entities will benefit the interoperability between BioNLP systems and associated applications. However, two major issues have been observed as hindering the optimal use of terminology/ontology for BioNLP: First, the existing terminology/ontology knowledge bases are not customized for NLP purposes and contain problematic contents; Second, automated solutions for improving and using the knowledge bases are still inadequate and therefore limiting their use in BioNLP.
520 $a To address the issues, corresponding solutions were proposed in the dissertation both to improve terminology/ontology for BioNLP purposes and to demonstrate feasibility of using terminology/ontology in BioNLP applications. For the first task, two automatic classifiers were developed to reclassify and audit semantic classification of terminology concepts. The classifiers use empirical language features and complement other auditing methods that apply ontological principles. For the second task, we developed unsupervised methods that use terminology/ontology for word sense disambiguation (WSD). The methods can help reduce the labor of manual annotation and sample representative evaluation instances for WSD research. Promising results have been achieved in both tasks and we have made the reclassified concepts a public database for the community. The results also enhanced our understanding about the biomedical terminology/ontology knowledge bases and pointed out interesting directions for future research. The methods by the dissertation can be generalized to other fields and should promote the use of standardized terminology/ontology in biomedicine and healthcare.
590 $a School code: 0054.
650 4 $a Biology, Bioinformatics. $3 1018415
650 4 $a Artificial Intelligence. $3 769149
690 $a 0715
690 $a 0800
710 2 $a Columbia University. $3 571054
773 0 $t Dissertation Abstracts International $g 71-03B.
790 1 0 $a Friedman, Carol, $e advisor
790 $a 0054
791 $a Ph.D.
792 $a 2009
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3400568