NLP Tech Talk for QBRC at UTSW

1. Q B R C T E C H T A L K
T H O M A S N A T E P E R S O N
C L I N I C A L S C I E N C E S
P R O S P R C E N T E R A N D Q B R C
M A Y 6 , 2 0 1 3
1
NLP: Natural Language Processing

2. 2
Outline
 Basics of NLP
 NLP Toolkits
 Basic Implementation Example
 Questions?

3. 3
What is NLP?
 Not:
 Natural Language Programming (NLP)
 Neuro-Linguistic Programing (NLP)
 “Natural Language Processing (NLP) is a field of
computer science, artificial intelligence, and
linguistics concerned with the interactions between
computers and human(natural) languages.”
-Wikipedia

4. 4
What can’t it do?
 Extract information not understandable or
discernible by “you”.
 Extract deeper meaning.
 Is not a substitute for Regular Expression pattern
matching

5. 5
Basics of NLP
 Large research field
 From Speech Recognitions to Optical Character Recognition
 Examples:
 Watson (Jeopardy)
 Cleverbot
 Siri/Dragon Speak
 Captcha
 I am only concerned about Information Extraction (IE)
 Sentence detection
 Part of Speech (POS) tagging
 (nouns, verbs, adverbs)
 Named-entity recognition (NER)
 (names, organizations, locations)
 Lemmatisation
 (Walk, walked, walks, walking)
 Relationship extraction
 All possible word relationships
 Parsing
 Determining most probable word relationships
 Coreference
 Linking of references between multiple sentences

6. 6
What’s the point of all that?
 Help categorize unstructured text into a more
structured format so that discrete information can
more easily be extracted.

7. 7
NLP Information Extraction Example
“Pierre Vinken, 61 years old, will join the board as a
nonexecutive director Nov. 29.”

8. 8
NLP Information Extraction Example
POS (Part of Speech) Tagging
 “Pierre Vinken, 61 years old, will join the board as a nonexecutive director Nov. 29.”
Pierre/NNP
Vinken/NNP
,/,
61/CD
years/NNS
old/JJ
,/,
will/MD
join/VB
the/DT
board/NN
as/IN
a/DT
nonexecutive/JJ
director/NN
Nov./NNP
29/CD
./.

9. Penn Treebank Tagset
 CC Coordinating conjunction e.g. and,but,or...
 CD Cardinal Number
 DT Determiner
 EX Existential there
 FW Foreign Word
 IN Preposision or subordinating conjunction
 JJ Adjective
 JJR Adjective, comparative
 JJS Adjective, superlative
 LS List Item Marker
 MD Modal e.g. can, could, might, may...
 NN Noun, singular or mass
 NNP Proper Noun, singular
 NNPS Proper Noun, plural
 NNS Noun, plural
 PDT Predeterminer e.g. all, both ... when they
precede an article
 POS Possessive Ending e.g. Nouns ending in 's
 PRP Personal Pronoun e.g. I, me, you, he...
 PRP$ Possessive Pronoun e.g. my, your, mine,
yours...
 RB Adverb Most words that end in -ly as well
as degree words like quite, too and very
 RBR Adverb, comparative Adverbs with the
comparative ending -er, with a strictly
comparative meaning.
 RBS Adverb, superlative
 RP Particle
 SYM Symbol Should be used for mathematical,
scientific or technical symbols
 TO to
 UH Interjection e.g. uh, well, yes, my...
 VB Verb, base form subsumes imperatives,
infinitives and subjunctives
 VBD Verb, past tense includes the conditional
form of the verb to be
 VBG Verb, gerund or persent participle
 VBN Verb, past participle
 VBP Verb, non-3rd person singular present
 VBZ Verb, 3rd person singular present
 WDT Wh-determiner e.g. which, and that when it
is used as a relative pronoun
 WP Wh-pronoun e.g. what, who, whom...
 WP$ Possessive wh-pronoun e.g.
 WRB Wh-adverb e.g. how, where why

10. 10
POS Parse Tree

11. 11
POS Parse Tree
“Pierre Vinken, 61 years old, will join the board as a nonexecutive director
Nov. 29.”
( (S (NP-SBJ (NP (NNP Pierre) (NNP Vinken))
(, ,)
(ADJP (NML (CD 61) (NNS years))
(JJ old))
(, ,))
(VP (MD will)
(VP (VB join)
(NP (DT the) (NN board))
(PP-CLR (IN as)
(NP (DT a) (JJ nonexecutive) (NN director)))
(NP-TMP (NNP Nov.) (CD 29))))
(. .)))

12. 12
NLP Information Extraction Example
POS (Part of Speech) Tagging
 “Pierre Vinken, 61 years old, will join the board as a nonexecutive director Nov. 29.”
Pierre/NNP
Vinken/NNP
,/,
61/CD
years/NNS
old/JJ
,/,
will/MD
join/VB
the/DT
board/NN
as/IN
a/DT
nonexecutive/JJ
director/NN
Nov./NNP
29/CD
./.

13. 13
NLP Information Extraction Example
NER (Named Entity Recognition) Tagging
 “Pierre Vinken, 61 years old, will join the board as a nonexecutive director Nov. 29.”
Pierre/NNP/PERSON
Vinken/NNP/PERSON
,/,/O
61/CD/DURATION
years/NNS/NUMBER
old/JJ/DURATION
,/,/O
will/MD/O
join/VB/O
the/DT/O
board/NN/O
as/IN/O
a/DT/O
nonexecutive/JJ/O
director/NN/O
Nov./NNP/DATE
29/CD/DATE
././O

14. 14
NLP Information Extraction Example
Lemmatisation
 “Pierre Vinken, 61 years old, will join the board as a nonexecutive director Nov. 29.”
Pierre/NNP/PERSON [Pierre]
Vinken/NNP/PERSON [Vinken]
,/,/O [,]
61/CD/DURATION [61]
years/NNS/NUMBER [year]
old/JJ/DURATION [old]
,/,/O [,]
will/MD/O [will]
join/VB/O [join]
the/DT/O [the]
board/NN/O [board]
as/IN/O [as]
a/DT/O [a]
nonexecutive/JJ/O [nonexecutive]
director/NN/O [director]
Nov./NNP/DATE [Nov.]
29/CD/DATE [29]
././O [.]

15. 15
Relationship Parsing
 “Pierre Vinken, 61 years old, will join the board as a nonexecutive
director Nov. 29.”
nn(Vinken-1, Pierre-0) [nn modifier]
nsubj(join-8, Vinken-1) [nominal subject]
num(years-4, 61-3) [numeric modifier]
npadvmod(old-5, years-4) [noun phrase adverbial modifier]
amod(Vinken-1, old-5) [adjectival modifier]
aux(join-8, will-7) [auxiliary]
det(board-10, the-9) [determiner]
dobj(join-8, board-10) [direct object]
det(director-14, a-12) [determiner]
amod(director-14, nonexecutive-13) [adjectival modifier]
prep_as(join-8, director-14) [prep_collapsed]
tmod(join-8, Nov.-15) [temporal modifier]
num(Nov.-15, 29-16) [numeric modifier]

16. Relationship Extraction
 root - root
 dep - dependent
 aux - auxiliary
 auxpass - passive auxiliary
 cop - copula
 arg - argument
 agent - agent
 comp - complement
 acomp - adjectival complement
 attr - attributive
 ccomp - clausal complement with internal
subject
 xcomp - clausal complement with external
subject
 complm - complementizer
 obj - object
• dobj - direct object
• iobj - indirect object
• pobj - object of preposition
 mark - marker (word introducing an advcl )
 rel - relative (word introducing a rcmod )
 subj - subject
 nsubj - nominal subject
• nsubjpass - passive nominal subject
 csubj - clausal subject
• csubjpass - passive clausal subject
 cc - coordination
 conj - conjunct
 expl - expletive (expletive “there”)
 mod - modifier
 abbrev - abbreviation modifier
 amod - adjectival modifier
 appos - appositional modifier
 advcl - adverbial clause modifier
 purpcl - purpose clause modifier
 det - determiner
 predet - predeterminer
 preconj - preconjunct
 infmod - infinitival modifier
 mwe - multi-word expression modifier
 partmod - participial modifier
 advmod - adverbial modifier
 neg - negation modifier
 rcmod - relative clause modifier
 quantmod - quantifier modifier
 nn - noun compound modifier
 npadvmod - noun phrase adverbial modifier
 tmod - temporal modifier
 num - numeric modifier
 number - element of compound number
 prep - prepositional modifier
 poss - possession modifier
 possessive - possessive modifier (’s)
 prt - phrasal verb particle
 parataxis - parataxis
 punct - punctuation
 ref - referent
 sdep - semantic dependent
 xsubj - controlling subject

17. 17
NLP Toolkits
 41 different toolkits listed in Wikipedia
 Four of the more popular free open source (FOSS) IE
toolkits

Name Language License Creators
OpenNLP Java
Apache
License
2.0
Online
community
General Architecture for
Text Engineering
(GATE)
Java LGPL
GATE open
source community
Natural Language
Toolkit (NLTK)
Python
Apache
2.0
Team NLTK
Stanford NLP Java GPL
The Stanford
Natural Language
Processing Group

18. 18
NLP Toolkits
 OpenNLP
 Extensive publications
 Corporate Sponsorship
 Java

19. 19
NLP Toolkits
 General Architecture for Text Engineering (GATE)
 Extensive publications
 Integrated Development Environment (IDE) to assist in
development
 Java
 Java Annotation Patterns Engine (JAPE)

20. 20
NLP Toolkits
 Natural Language Tool Kit (NLTK)
 Extensive publications
 Two published documentation books from O’Reilly and Packt

21. 21
NLP Toolkits
 Stanford Core NLP
 Extensive publications
 Wrappers for Perl, Python, Ruby, and Scala languages
 Plugins for GATE and NLTK

22. 22
Questions from PROSPR to answer
 From the hand typed Colonoscopy report:
 How many Polyps
 Location of Polyps
 Size of Polyps

23. 23
Sample Workflow
 Report Definition
 Report Sectionization
 Formatting the Text
 Process the Section
 Further analysis

24. Report Example
Gastroenterology Laboratory
Patient Name: Susan Storm Richards
Procedure Date: 5/06/2013 15:00:15 PM
MRN: 123456789
Age: 60
Accession #: 123456
Gender: Female
Order #: 123456789
Ethnicity:
Attending MD: Victor Von Doom MD
Note Status: Finalized
Room: 666
Procedure: Colonoscopy
Referring MD: Reed Richards
Providers: Victor von Doom, MD (Doctor)
Attending Participation: I personally performed the entire procedure.
Medicines: SomeDrug 3 mg IV, OtherDrug 75 micrograms IV
Indications: Screening for colorectal malignant neoplasm
Complications: No immediate complications.
Patient Profile: Refer to note in patient chart for documentation of history and
physical.
Procedure: Pre-Anesthesia Assessment:
- PLorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris
nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia
deserunt mollit anim id est laborum. ASA Grade Assessment: II - Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim
veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint
occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
Lorem ipsum dolor sit amet, consectetur adipisicing elit,
sed do eiusmod tempor incididunt ut labore et dolore magna
aliqua. Ut enim ad minim veniam, quis nostrud exercitation
ullamco laboris nisi ut aliquip ex ea commodo consequat.
Duis aute irure dolor in reprehenderit in
esse cillum dolore eu fugiat nulla pariatur. Excepteur sint
occaecat cupidatat non proident, sunt in culpa qui officia
deserunt mollit anim id est laborum.
Findings: Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud
exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu f ugiat nulla
pariatur. Three pedunculated polyps were found in the mid sigmoid colon and in the proximal ascending colonThe polyps were 30 mm in size. Excepteur sint occaecat cupidatat non proident,
sunt in culpa qui officia deserunt mollit anim id est laborum.
Estimated Blood Loss: Estimated blood loss: none.
Recommendation: - Discharge patient to home (ambulatory).
- High fiber diet indefinitely.
CPT(c) Code(s): --- Technical ---
G0121, Colorectal cancer screening; colonoscopy on individual
not meeting criteria for high risk
CPT Copyright 2010 American Medical Association. All Rights Reserved.
The codes documented in this report are preliminary and upon coder review may be revised
to meet current compliance requirements.
Victor von Doom
Victor von Doom, MD
5/6/2013 15:10
This report has been signed electronically.
Number of Addenda: 0

25. 25
Sectioned
Findings: Lorem ipsum dolor sit amet, consectetur
adipisicing elit, sed do eiusmod tempor incididunt ut
labore et dolore magna aliqua. Ut enim ad minim veniam,
quis nostrud exercitation ullamco laboris nisi ut aliquip ex
ea commodo consequat. Duis aute irure dolor in
reprehenderit in voluptate velit esse cillum dolore eu fugiat
nulla pariatur. Three pedunculated polyps were
found in the mid sigmoid colon and in the
proximal ascending colonThe polyps were 30 mm
in size. Excepteur sint occaecat cupidatat non
proident, sunt in culpa qui officia deserunt mollit anim id
est laborum.

26. 26
Sample
“ Three pedunculated polyps were found in the mid
sigmoid colon and in the proximal ascending
colonThe polyps were 30 mm in size. ”

27. 27
Regex Formatting Text: Removing Spaces
$text =~s/(bd+b)(s)(bmmb)/$1$3/g;
$text =~s/(b[a-z]+)([A-Z])([a-z]+b)/$1.s$2$3/g;
$text =~ s/^s+//;
$text =~ s/s+$//;

28. 28
Formatted Sample
“Three pedunculated polyps were found in the mid
sigmoid colon and in the proximal ascending colon.
The polyps were 30mm in size.”

29. NLP Information Extraction Example
Relationship Dependencies
 Original sentence:
Three pedunculated polyps were found in the mid
sigmoid colon and in the proximal ascending colon.
Dependencies:
num(polyps-2, Three-0) [numeric modifier]
amod(polyps-2, pedunculated-1) [adjectival
modifier]
nsubjpass(found-4, polyps-2) [nominal passive
subject]
auxpass(found-4, were-3) [passive auxiliary]
det(colon-9, the-6) [determiner]
amod(colon-9, mid-7) [adjectival modifier]
nn(colon-9, sigmoid-8) [nn modifier]
prep_in(found-4, colon-9) [prep_collapsed]
det(proximal-13, the-12) [determiner]
prep_in(found-4, proximal-13) [prep_collapsed]
conj_and(colon-9, proximal-13) [conj_collapsed]
partmod(proximal-13, ascending-14) [participial
modifier]
dobj(ascending-14, colon-15) [direct object]
 Original sentence:
The polyps were 30mm in size.
Dependencies:
det(polyps-1, The-0) [determiner]
nsubj(30mm-3, polyps-1) [nominal subject]
cop(30mm-3, were-2) [copula]
prep_in(30mm-3, size-5) [prep_collapsed]

30. 30
Output
“Three pedunculated polyps were found in the mid
sigmoid colon and in the proximal ascending colon.
The polyps were 30 mm in size.”
Output
Number of Polyps: 3
Size of Polyps: 30,
Location of Polyps: 1,4,

31. 1 use Lingua::StanfordCoreNLP;
2 use Lingua::EN::Words2Nums;
3 use strict;
4 use warnings;
5 my $pipeline = new Lingua::StanfordCoreNLP::Pipeline(1);
6 my $text = "Three pedunculated polyps were found in the mid sigmoid colon and in the proximal ascending colonThe polyps were 30 mm in size.";
7 $text =~s/(bd+b)(s)(bmmb)/$1$3/g;
8 $text =~s/(b[a-z]+)([A-Z])([a-z]+b)/$1.s$2$3/g;
9 $text =~ s/^s+//;
10 $text =~ s/s+$//;
11 my $result = $pipeline->process($text);
12 my $polypCount;
13 my $polypSize;
14 my $polypLocation;
15 for my $sentence (@{$result->toArray})
16 {
17 for my $dep (@{$sentence->getDependencies->toArray})
18 {
19 my $relation = $dep->getRelation,
20 my $govern = $dep->getGovernor->getWord,
21 my $depend = $dep->getDependent->getWord;
22 my $num=words2nums($depend);
23
24 if(($relation eq "num")&&($govern=~/^polyp(|s)$/i))
25 {
26 $polypCount=$num;
27 }
28 if(($relation eq "nsubj")&&($govern=~/^d+mm$/)&&($depend=~/^polyp(|s)$/i))
29 {
30 $govern=~s/mm$//;
31 $polypSize="$govern,";
32 }
33 if(($relation eq "nn")&&($govern=~/^colon$/i)&&($depend=~/sigmoid/i))
34 {
35 $polypLocation="1,";
36 }
37 if(($relation eq "dobj")&&($govern=~/^ascending$/i)&&($depend=~/^colon$/i))
38 {
39 $polypLocation.="4,";
40 }
41 }
42 }
43 print "Number of Polyps:t$polypCountn";
44 print "Size of Polyps:tt$polypSizen";
45 print "Location of Polyps:t$polypLocationn";
Perl Example

32. 32
F - Score
6/26/2013
 Comparison against a manually curated “Gold
Standard”
 Precision = Proportion of True Positives
 Recall = True Proportion of Actual Positives

33. 33
Questions?!