Deriving Consensus for Multi-Parallel Corpora: An English Bible - - PowerPoint PPT Presentation

deriving consensus for multi parallel corpora an english
SMART_READER_LITE
LIVE PREVIEW

Deriving Consensus for Multi-Parallel Corpora: An English Bible - - PowerPoint PPT Presentation

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study Patrick Xia David Yarowsky code, resources, slides: www.github.com/pitrack/monolign Deriving Consensus for Multi-Parallel Corpora: An English Bible Study Deriving


slide-1
SLIDE 1

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Patrick Xia David Yarowsky

code, resources, slides: www.github.com/pitrack/monolign

slide-2
SLIDE 2

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

slide-3
SLIDE 3

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

slide-4
SLIDE 4

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

I like Taipei

slide-5
SLIDE 5

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

I like Taipei 我喜欢台北

slide-6
SLIDE 6

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

I like Taipei 我喜欢台北

slide-7
SLIDE 7

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

I like Taipei 我喜欢台北

N:1sg V N:Loc

slide-8
SLIDE 8

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext

I like Taipei 我喜欢台北

N:1sg V N:Loc N:1sg V N:Loc

slide-9
SLIDE 9

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北

N:1sg V N:Loc N:1sg V N:Loc

slide-10
SLIDE 10

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北

N:1sg V N:Loc N:1sg V N:Loc

slide-11
SLIDE 11

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北

N:1sg V N:Loc N:1sg V N:Loc

slide-12
SLIDE 12

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北

N:1sg V N:Loc N:1sg V N:Loc

slide-13
SLIDE 13

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北 I love Taipei

N:1sg V N:Loc N:1sg V N:Loc

slide-14
SLIDE 14

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北 I love Taipei

N:1sg V N:Loc N:1sg V N:Loc

slide-15
SLIDE 15

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北 I love Taipei

N:1sg V N:Loc N:1sg V N:Loc N:1sg V N:Loc

slide-16
SLIDE 16

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北 I love Taipei

N:1sg V N:Loc N:1sg V N:Loc N:1sg V N:Loc N:1sg V N:Loc

slide-17
SLIDE 17

I like Taiwan

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

Bitext Multi-parallel corpora

I like Taipei 我喜欢台北 我喜欢台北 我♡台北 I love Taipei

N:1sg V N:Loc N:1sg V N:Loc N:1sg V N:Loc N:1sg V N:Loc

“Tag Projection”

slide-18
SLIDE 18

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

(Mayer and Cysouw, 2014)

slide-19
SLIDE 19

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

  • 800+ languages, verse-aligned
  • 27 English versions (New Testament)

(Mayer and Cysouw, 2014)

slide-20
SLIDE 20

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

  • 800+ languages, verse-aligned
  • 27 English versions (New Testament)
  • Improved English consensus → improved low-

resource tag projection from a common source

(Mayer and Cysouw, 2014)

slide-21
SLIDE 21

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

  • 800+ languages, verse-aligned
  • 27 English versions (New Testament)
  • Improved English consensus → improved low-

resource tag projection from a common source

(Mayer and Cysouw, 2014)

New Simplified: Montgomery: Lexham: Then Herod secretly called the astrologers … Thereupon Herod sent secretly for the Magi … Then Herod secretly summoned the wise men …

Examples:

slide-22
SLIDE 22

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

slide-23
SLIDE 23

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Then Herod secretly called the astrologers …

slide-24
SLIDE 24

Thereupon Herod sent secretly for the Magi …

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Then Herod secretly called the astrologers …

slide-25
SLIDE 25

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Thereupon Herod sent secretly for the Magi … Then Herod secretly called the astrologers …

slide-26
SLIDE 26

Then Herod secretly summoned the wise men …

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Lexham Thereupon Herod sent secretly for the Magi … Then Herod secretly called the astrologers …

slide-27
SLIDE 27

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Lexham Thereupon Herod sent secretly for the Magi … Then Herod secretly called the astrologers … Then Herod secretly summoned the wise men …

slide-28
SLIDE 28

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Lexham Thereupon Herod sent secretly for the Magi … Then Herod secretly called the astrologers … Then Herod secretly summoned the wise men …

Where do these go if the goal is: Paraphrases? Tag projection?

slide-29
SLIDE 29

Deriving Consensus for Multi-Parallel Corpora: An English Bible Study

New Simplified: Montgomery: Lexham Thereupon Herod sent secretly for the Magi … Then Herod secretly called the astrologers … Then Herod secretly summoned the wise men …

Where do these go if the goal is: Paraphrases? Tag projection?

Which one would a low- resource Bible likely choose?

slide-30
SLIDE 30

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matchings
  • 3. Consensus + Resources

New Simplified: Montgomery: Lexham: Then Herod secretly called the astrologers … Thereupon Herod sent secretly for the Magi … Then Herod secretly summoned the wise men …

slide-31
SLIDE 31

Creating a candidate matching

  • Randomly select first document

Then Herod secretly called the astrologers New Simplified:

slide-32
SLIDE 32

Creating a candidate matching

  • Randomly select first document

Then Herod secretly called the astrologers RB NNP RB VBD DT NNS New Simplified:

slide-33
SLIDE 33

Creating a candidate matching

  • Initialize matching

Then Herod secretly called the astrologers RB NNP RB VBD DT NNS Matching:

slide-34
SLIDE 34

Creating a candidate matching

  • Initialize matching

Score: 0 New Simplified

Then Herod secretly called the astrologers RB NNP RB VBD DT NNS Matching:

slide-35
SLIDE 35
  • For each document, score possible edges

Thereupon Herod sent secretly for Magi the Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-36
SLIDE 36
  • For each document, score possible edges

Thereupon Herod sent secretly for Magi the RB NNP RB VBD DT NNS IN Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-37
SLIDE 37
  • For each document, score possible edges

Thereupon Herod sent secretly for Magi the RB NNP RB VBD DT NNS IN Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-38
SLIDE 38
  • For each document, score possible edges

Thereupon Herod sent secretly for Magi the RB NNP RB VBD DT NNS IN Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-39
SLIDE 39
  • Take maximum weight matching

Thereupon Herod sent secretly for Magi the RB NNP RB VBD DT NNS IN Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-40
SLIDE 40
  • Take maximum weight matching

Thereupon Herod sent secretly for Magi the RB NNP RB VBD DT NNS IN Then Herod secretly called the astrologers RB NNP RB VBD DT NNS

Score: 0 New Simplified →Montgomery

Matching: Montgomery:

Creating a candidate matching

slide-41
SLIDE 41

Creating a candidate matching

  • Merge

Matching:

Then, Thereupon Herod secretly called, sent for astrologers, Magi

RB NNP RB VBD DT NNS IN

the

slide-42
SLIDE 42

Creating a candidate matching

  • Merge

Matching:

Score: 20 New Simplified Montgomery

Then, Thereupon Herod secretly called, sent for astrologers, Magi

RB NNP RB VBD DT NNS IN

the

slide-43
SLIDE 43

Creating a candidate matching

  • Repeat for each document

Matching: Lexham: Then Herod secretly summoned the wise men

Then, Thereupon Herod secretly called, sent astrologers, Magi

RB NNP RB VBD DT NNS IN

Score: 20 New Simplified Montgomery →Lexham

for the

slide-44
SLIDE 44

Creating a candidate matching

  • Repeat for each document

Matching: Lexham: Then Herod secretly summoned the wise men RB NNP RB VBD DT NNS JJ

Then, Thereupon Herod secretly called, sent astrologers, Magi

RB NNP RB VBD DT NNS IN

Score: 20 New Simplified Montgomery →Lexham

for the

slide-45
SLIDE 45

Creating a candidate matching

  • Repeat for each document

Matching: Lexham: Then Herod secretly summoned the wise men RB NNP RB VBD DT NNS JJ

Then, Thereupon Herod secretly called, sent astrologers, Magi

RB NNP RB VBD DT NNS IN

Score: 20 New Simplified Montgomery →Lexham

for the

slide-46
SLIDE 46

Creating a candidate matching

  • Repeat for each document

Matching: Lexham: Then Herod secretly summoned the wise men RB NNP RB VBD DT NNS JJ

Then, Thereupon Herod secretly called, sent astrologers, Magi

RB NNP RB VBD DT NNS IN

Score: 20 New Simplified Montgomery →Lexham

for the

slide-47
SLIDE 47

Creating a candidate matching

  • Repeat for each document

Matching: Lexham: Then Herod secretly summoned the wise men RB NNP RB VBD DT NNS JJ

Then, Thereupon Herod secretly called, sent astrologers, Magi

RB NNP RB VBD DT NNS IN

Score: 20 New Simplified Montgomery →Lexham

for the

slide-48
SLIDE 48

Creating a candidate matching

Matching:

Then, Thereupon Herod secretly called, sent, summoned for astrologers, Magi, men

VBD

the wise

RB NNP RB DT NNS IN JJ

Score: 40 New Simplified Montgomery Lexham

slide-49
SLIDE 49

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matchings
  • 3. Consensus + Resources

Score: 40 New Simplified Montgomery Lexham

slide-50
SLIDE 50

Corpus Matchings

slide-51
SLIDE 51

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order
slide-52
SLIDE 52

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order

Score: 40 New Simplified Montgomery Lexham Iteration 1

slide-53
SLIDE 53

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order

Score: 40 Score: 45 New Simplified Montgomery Lexham New Simplified Lexham Montgomery Iteration 1 Iteration 2

slide-54
SLIDE 54

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order

Score: 40 Score: 45 Score: 38 New Simplified Montgomery Lexham New Simplified Lexham Montgomery Lexham Montgomery New Simplified Iteration 1 Iteration 2 Iteration 3

slide-55
SLIDE 55

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order

Score: 40 Score: 45 Score: 38 Score: 42 New Simplified Montgomery Lexham New Simplified Lexham Montgomery Lexham Montgomery New Simplified Lexham New Simplified Montgomery Iteration 1 Iteration 2 Iteration 3 Iteration 4

slide-56
SLIDE 56

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order
  • For each verse, keep the maximum weight

matching

Score: 40 Score: 45 Score: 38 Score: 42 New Simplified Montgomery Lexham New Simplified Lexham Montgomery Lexham Montgomery New Simplified Lexham New Simplified Montgomery Iteration 1 Iteration 2 Iteration 3 Iteration 4

slide-57
SLIDE 57

Corpus Matchings

  • Run for N=10 iterations
  • Rescore edge weights
  • Re-align bitexts with inferred word lists
  • Randomization of document order
  • For each verse, keep the maximum weight

matching

Score: 40 Score: 45 Score: 38 Score: 42 New Simplified Montgomery Lexham New Simplified Lexham Montgomery Lexham Montgomery New Simplified Lexham New Simplified Montgomery Iteration 1 Iteration 2 Iteration 3 Iteration 4

slide-58
SLIDE 58

Corpus Matchings

Then, Thereupon, first Herod, Herodes secretly, privately called, sent, summoned for astrologers, Magi, men, Magians, magi

VBD

the wise

RB NNP RB DT NNS IN JJ Best matching:

slide-59
SLIDE 59

Corpus Matchings

  • A Bible in a low-resource language can align or project to a single matching

instead of several different versions

Then, Thereupon, first Herod, Herodes secretly, privately called, sent, summoned for astrologers, Magi, men, Magians, magi

VBD

the wise

RB NNP RB DT NNS IN JJ Best matching:

slide-60
SLIDE 60

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources
slide-61
SLIDE 61

Consensus + Resources

  • Another format:

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for for the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men magi men

slide-62
SLIDE 62

Consensus + Resources

  • Another format:

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for for the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men magi men

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for For the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men Magi men

slide-63
SLIDE 63

Word Tag Then RB Herod NNP , . secretly RB called VBD for IN the DT

wise JJ men NNS

Consensus + Resources

  • Another format:

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for for the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men magi men

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for For the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men Magi men

slide-64
SLIDE 64

Word Tag Then RB Herod NNP , . secretly RB called VBD for IN the DT

wise JJ men NNS

Consensus + Resources

  • Another format:

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for for the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men magi men

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for For the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men Magi men

slide-65
SLIDE 65

Word Tag Then RB Herod NNP , . secretly RB called VBD for IN the DT

wise JJ men NNS

Consensus + Resources

  • Another format:

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for for the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men magi men

worldwide newsimplified montgomery etheridge godword majority lexham common contemporary

Then Then Thereupon Then Then Then Then Then first Herod Herod Herod Herod Herodes Herod Herod Herod Herod , secretly secretly secretly privately secretly secretly secretly secretly secretly called called sent called called called summoned called called for For the the the the the the the the the wise wise wise wise wise men astrologers Magi Magians men men men Magi men

Consensus

slide-66
SLIDE 66

Consensus + Resources

slide-67
SLIDE 67

Consensus + Resources

Consensus POS and dependency parses reinforce annotations

slide-68
SLIDE 68

Consensus + Resources

Consensus POS and dependency parses reinforce annotations

  • Softer Consensus Tags
slide-69
SLIDE 69

Consensus + Resources

Consensus POS and dependency parses reinforce annotations

  • Softer Consensus Tags
  • POS:
  • “TIME”: Sg. Noun (1.00) → Sg. Noun (0.94), Pl. Noun (0.05) …
  • “SECRET”: Sg. Noun (0.54), Adj (0.46) → Adj (0.51), Sg. Noun (0.47) …
slide-70
SLIDE 70

Consensus + Resources

Consensus POS and dependency parses reinforce annotations

  • Softer Consensus Tags
  • POS:
  • “TIME”: Sg. Noun (1.00) → Sg. Noun (0.94), Pl. Noun (0.05) …
  • “SECRET”: Sg. Noun (0.54), Adj (0.46) → Adj (0.51), Sg. Noun (0.47) …
  • Lexical Heads:
  • “SECRET”: in (0.28), is (0.06), kept (0.04) … → in (0.32), place (0.05), mystery (0.04)…
slide-71
SLIDE 71

Consensus + Resources

Consensus POS and dependency parses reinforce annotations

  • Softer Consensus Tags
  • POS:
  • “TIME”: Sg. Noun (1.00) → Sg. Noun (0.94), Pl. Noun (0.05) …
  • “SECRET”: Sg. Noun (0.54), Adj (0.46) → Adj (0.51), Sg. Noun (0.47) …
  • Lexical Heads:
  • “SECRET”: in (0.28), is (0.06), kept (0.04) … → in (0.32), place (0.05), mystery (0.04)…

Paraphrase Discovery

  • “SECRET”: secret (0.59), mystery (0.19), private (0.04) …
slide-72
SLIDE 72

Consensus + Resources

slide-73
SLIDE 73

Consensus + Resources

Richer exploration and analysis of text meaning

slide-74
SLIDE 74

Consensus + Resources

Richer exploration and analysis of text meaning

  • Translator/author variation:
  • “HYMENAEUS”: Hymenaues (0.82), Hymenius (0.04), Hymeneus (0.04), Humenaios (0.04)

slide-75
SLIDE 75

Consensus + Resources

Richer exploration and analysis of text meaning

  • Translator/author variation:
  • “HYMENAEUS”: Hymenaues (0.82), Hymenius (0.04), Hymeneus (0.04), Humenaios (0.04)

  • Historical language usage:
  • “BLAZES”: burns (0.48), burning (0.33), burneth (0.10), blazes (0.04)
slide-76
SLIDE 76

Consensus + Resources

Richer exploration and analysis of text meaning

  • Translator/author variation:
  • “HYMENAEUS”: Hymenaues (0.82), Hymenius (0.04), Hymeneus (0.04), Humenaios (0.04)

  • Historical language usage:
  • “BLAZES”: burns (0.48), burning (0.33), burneth (0.10), blazes (0.04)
  • In-domain word choices and disambiguation:
  • “CHALLENGED”: said (0.15), opposed (0.13), urged (0.12), tested (0.10), tempted (0.06) …
slide-77
SLIDE 77

Consensus + Resources

Richer exploration and analysis of text meaning

  • Translator/author variation:
  • “HYMENAEUS”: Hymenaues (0.82), Hymenius (0.04), Hymeneus (0.04), Humenaios (0.04)

  • Historical language usage:
  • “BLAZES”: burns (0.48), burning (0.33), burneth (0.10), blazes (0.04)
  • In-domain word choices and disambiguation:
  • “CHALLENGED”: said (0.15), opposed (0.13), urged (0.12), tested (0.10), tempted (0.06) …

Can shed insight into textual or theological debates

slide-78
SLIDE 78

Deriving a Consensus

slide-79
SLIDE 79

Deriving a Consensus

  • 1. Creating a candidate matching
slide-80
SLIDE 80

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
slide-81
SLIDE 81

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources
slide-82
SLIDE 82

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources

What’s next?

slide-83
SLIDE 83

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources

What’s next? Up to you! Code/resources/slides available at

slide-84
SLIDE 84

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources

What’s next? Up to you! Code/resources/slides available at

www.github.com/pitrack/monolign

slide-85
SLIDE 85

Deriving a Consensus

  • 1. Creating a candidate matching
  • 2. Corpus Matching
  • 3. Consensus + Resources

What’s next? Up to you! Code/resources/slides available at

www.github.com/pitrack/monolign Thank You!