Annotating patient clinical records with syntactic chunks and named entities: the Harvey Corpus

Aleksandar Savkov; John Carroll; Rob Koeling; Jackie Cassell

doi:10.1007/s10579-015-9330-7

Annotating patient clinical records with syntactic chunks and named entities: the Harvey Corpus

Aleksandar Savkov, John Carroll, Rob Koeling, Jackie Cassell

Research output: Contribution to journal › Article › peer-review

31 Scopus citations

Abstract

The free text notes typed by physicians during patient consultations contain valuable information for the study of disease and treatment. These notes are difficult to process by existing natural language analysis tools since they are highly telegraphic (omitting many words), and contain many spelling mistakes, inconsistencies in punctuation, and non-standard word order. To support information extraction and classification tasks over such text, we describe a de-identified corpus of free text notes, a shallow syntactic and named entity annotation scheme for this kind of text, and an approach to training domain specialists with no linguistic background to annotate the text. Finally, we present a statistical chunking system for such clinical text with a stable learning rate and good accuracy, indicating that the manual annotation is consistent and that the annotation scheme is tractable for machine learning.

Original language	English
Pages (from-to)	523-548
Number of pages	26
Journal	Language Resources and Evaluation
Volume	50
Issue number	3
DOIs	https://doi.org/10.1007/s10579-015-9330-7
State	Published - Sep 1 2016
Externally published	Yes

Keywords

Annotation guidelines
Chunking
Clinical text
Corpus annotation
Named entities

Access to Document

10.1007/s10579-015-9330-7

Cite this

@article{6e874c2a9f104223adf2a35ec92dc50b,

title = "Annotating patient clinical records with syntactic chunks and named entities: the Harvey Corpus",

abstract = "The free text notes typed by physicians during patient consultations contain valuable information for the study of disease and treatment. These notes are difficult to process by existing natural language analysis tools since they are highly telegraphic (omitting many words), and contain many spelling mistakes, inconsistencies in punctuation, and non-standard word order. To support information extraction and classification tasks over such text, we describe a de-identified corpus of free text notes, a shallow syntactic and named entity annotation scheme for this kind of text, and an approach to training domain specialists with no linguistic background to annotate the text. Finally, we present a statistical chunking system for such clinical text with a stable learning rate and good accuracy, indicating that the manual annotation is consistent and that the annotation scheme is tractable for machine learning.",

keywords = "Annotation guidelines, Chunking, Clinical text, Corpus annotation, Named entities",

author = "Aleksandar Savkov and John Carroll and Rob Koeling and Jackie Cassell",

note = "Publisher Copyright: {\textcopyright} 2016, The Author(s).",

year = "2016",

month = sep,

day = "1",

doi = "10.1007/s10579-015-9330-7",

language = "Ingl{\'e}s",

volume = "50",

pages = "523--548",

journal = "Language Resources and Evaluation",

issn = "1574-020X",

publisher = "Springer Netherlands",

number = "3",

}

TY - JOUR

T1 - Annotating patient clinical records with syntactic chunks and named entities

T2 - the Harvey Corpus

AU - Savkov, Aleksandar

AU - Carroll, John

AU - Koeling, Rob

AU - Cassell, Jackie

PY - 2016/9/1

Y1 - 2016/9/1

N2 - The free text notes typed by physicians during patient consultations contain valuable information for the study of disease and treatment. These notes are difficult to process by existing natural language analysis tools since they are highly telegraphic (omitting many words), and contain many spelling mistakes, inconsistencies in punctuation, and non-standard word order. To support information extraction and classification tasks over such text, we describe a de-identified corpus of free text notes, a shallow syntactic and named entity annotation scheme for this kind of text, and an approach to training domain specialists with no linguistic background to annotate the text. Finally, we present a statistical chunking system for such clinical text with a stable learning rate and good accuracy, indicating that the manual annotation is consistent and that the annotation scheme is tractable for machine learning.

AB - The free text notes typed by physicians during patient consultations contain valuable information for the study of disease and treatment. These notes are difficult to process by existing natural language analysis tools since they are highly telegraphic (omitting many words), and contain many spelling mistakes, inconsistencies in punctuation, and non-standard word order. To support information extraction and classification tasks over such text, we describe a de-identified corpus of free text notes, a shallow syntactic and named entity annotation scheme for this kind of text, and an approach to training domain specialists with no linguistic background to annotate the text. Finally, we present a statistical chunking system for such clinical text with a stable learning rate and good accuracy, indicating that the manual annotation is consistent and that the annotation scheme is tractable for machine learning.

KW - Annotation guidelines

KW - Chunking

KW - Clinical text

KW - Corpus annotation

KW - Named entities

UR - http://www.scopus.com/inward/record.url?scp=84954170687&partnerID=8YFLogxK

U2 - 10.1007/s10579-015-9330-7

DO - 10.1007/s10579-015-9330-7

M3 - Artículo

AN - SCOPUS:84954170687

SN - 1574-020X

VL - 50

SP - 523

EP - 548

JO - Language Resources and Evaluation

JF - Language Resources and Evaluation

IS - 3

ER -

Annotating patient clinical records with syntactic chunks and named entities: the Harvey Corpus

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this