phrasal rank encoding
play

Phrasal Rank-Encoding Exploiting Phrase Redundancy and - PowerPoint PPT Presentation

Jun 23, 2023 •290 likes •1.05k views

Phrasal Rank-Encoding Exploiting Phrase Redundancy and Translational Relations for Phrase Table Compression Marcin Junczys-Dowmunt Department of Mathematics and Computer Science Adam Mickiewicz University in Pozna n ul.Umultowska 87, 61-614

  1. Phrasal Rank-Encoding Exploiting Phrase Redundancy and Translational Relations for Phrase Table Compression Marcin Junczys-Dowmunt Department of Mathematics and Computer Science Adam Mickiewicz University in Pozna´ n ul.Umultowska 87, 61-614 Pozna´ n, Poland Global Databases Section World Intelectual Property Organizaton (WIPO) 34, chemin des Colombettes, 1211 Geneva, Switzerland junczys@amu.edu.pl September 4, 2012 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 1 / 31

  2. Translation Model Size 70 Phrase T. 60 30.4 + 27.6 Reord. T. 50 40 Gigabytes 30 20 10 0 Text M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 2 / 31

  3. Translation Model Size 70 Phrase T. 60 30.4 + 27.6 Reord. T. 50 40 Gigabytes 30 20 5.9 + 2.6 10 0 Text Gzipped M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 2 / 31

  4. Translation Model Size 70 22.0 + 44.4 Phrase T. 60 30.4 + 27.6 Reord. T. 50 40 Gigabytes 30 20 5.9 + 2.6 10 0 Text Gzipped Moses M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 2 / 31

  5. Translation Model Size 70 22.0 + 44.4 Phrase T. 60 30.4 + 27.6 Reord. T. 50 40 Gigabytes 30 20 5.9 + 2.6 10 4.8 + 1.4 0 Text Gzipped Moses Compact M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 2 / 31

  6. Translation Model Size 70 22.0 + 44.4 Phrase T. 60 30.4 + 27.6 Reord. T. 50 40 Gigabytes 30 20 5.9 + 2.6 10 4.8 + 1.4 2.9 + 1.4 0 Text Gzipped Moses Compact PR−Enc M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 2 / 31

  7. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 3 / 31

  8. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 3 / 31

  9. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch Maria no daba una bofetada a la bruja verde ||| Mary did not slap the green witch ||| 0-0 1-1 1-2 2-3 no daba una bofetada a la bruja verde ||| did not slap the green witch ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 Maria no daba una bofetada a la ||| Mary did not slap the ||| 0-0 1-1 1-2 2-3 3-3 4-3 5-4 6-4 daba una bofetada a la bruja verde ||| slap the green witch ||| 0-0 1-0 2-0 3-1 4-1 5-3 6-2 Maria no daba una bofetada ||| Mary did not slap ||| 0-0 1-1 1-2 2-3 3-3 4-3 no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 a la bruja verde ||| the green witch ||| 0-0 1-0 2-2 3-1 Maria no ||| Mary did not ||| 0-0 1-1 1-2 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 bruja verde ||| green witch ||| 0-1 1-0 Maria ||| Mary ||| 0-0 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 bruja ||| witch ||| 0-0 verde ||| green ||| 0-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 4 / 31

  10. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 5 / 31

  11. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 5 / 31

  12. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 5 / 31

  13. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  14. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  15. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  16. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) the ||| 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  17. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) the ||| 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  18. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) (1,0,0) ||| 4-3 5-3 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  19. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) (1,0,0) ||| no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  20. bofetada bruja Maria no daba una a la verde Mary did not slap the green witch no daba una bofetada a la ||| (0,2,0) (1,0,0) ||| no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 6 / 31

  21. Maria no daba una bofetada a la bruja verde ||| Mary did not slap the green witch ||| 0-0 1-1 1-2 2-3 no daba una bofetada a la bruja verde ||| did not slap the green witch ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 Maria no daba una bofetada a la ||| Mary did not slap the ||| 0-0 1-1 1-2 2-3 3-3 4-3 5-4 6-4 daba una bofetada a la bruja verde ||| slap the green witch ||| 0-0 1-0 2-0 3-1 4-1 5-3 6-2 Maria no daba una bofetada ||| Mary did not slap ||| 0-0 1-1 1-2 2-3 3-3 4-3 no daba una bofetada a la ||| did not slap the ||| 0-0 0-1 1-2 2-2 3-2 4-3 5-3 a la bruja verde ||| the green witch ||| 0-0 1-0 2-2 3-1 Maria no ||| Mary did not ||| 0-0 1-1 1-2 no daba una bofetada ||| did not slap ||| 0-0 0-1 1-2 2-2 3-2 daba una bofetada a la ||| slap the ||| 0-0 1-0 2-0 3-1 4-1 bruja verde ||| green witch ||| 0-1 1-0 Maria ||| Mary ||| 0-0 no ||| did not ||| 0-0 0-1 daba una bofetada ||| slap ||| 0-0 1-0 2-0 a la ||| the ||| 0-0 1-0 bruja ||| witch ||| 0-0 verde ||| green ||| 0-0 M. Junczys-Dowmunt (UAM) Phrasal Rank-Encoding September 2012 7 / 31

Recommend

2 3 4 5 8 9 MINNEAPOLIS MILWAUKEE MSA RANK #16 MSA RANK #39 CHICAGO MSA RANK #3

2 3 4 5 8 9 MINNEAPOLIS MILWAUKEE MSA RANK #16 MSA RANK #39 CHICAGO MSA RANK #3

2 3 4 5 8 9 MINNEAPOLIS MILWAUKEE MSA RANK #16 MSA RANK #39 CHICAGO MSA RANK #3 INDIANAPOLIS MSA RANK #34 DETROIT MSA RANK #14 COLUMBUS MSA RANK #33 BALTIMORE/ DC CORRIDOR MSA RANK #21 CINCINNATI MSA RANK #28 ATLANTA MSA RANK

726 views • 44 slides
Non-Headed Structures and Phrasal Constructions Jackendoff (2011) gives the following examples

Non-Headed Structures and Phrasal Constructions Jackendoff (2011) gives the following examples

Unifying Everything: Simpler Syntax, Construction Grammar, Minimalism and HPSG Non-Headed Structures and Phrasal Constructions Non-Headed Structures and Phrasal Constructions Jackendoff (2011) gives the following examples for phrasal

119 views • 9 slides
Advanced Lesson 26 Topic 26: Phrasal verbs literal and idiomatic meaning A phrasal verb is a

Advanced Lesson 26 Topic 26: Phrasal verbs literal and idiomatic meaning A phrasal verb is a

2 nd semester Advanced Lesson 26 Topic 26: Phrasal verbs literal and idiomatic meaning A phrasal verb is a verb with one or several prepositions. Unlike verbs with dependent prepositions, they have a completely new meaning. Literal phrasal

340 views • 6 slides
ITG for Joint Phrasal Translation Modeling Colin Cherry Dekang Lin University of Alberta

ITG for Joint Phrasal Translation Modeling Colin Cherry Dekang Lin University of Alberta

ITG for Joint Phrasal Translation Modeling Colin Cherry Dekang Lin University of Alberta Google Inc. April 26, 2007 University of Alberta The Gist Joint phrasal translation models (JPTM) learn a bilingual phrase table using EM

341 views • 31 slides
Prominence-based licensing in head movement and phrasal movement Brian Hsu LSA 2020 Annual

Prominence-based licensing in head movement and phrasal movement Brian Hsu LSA 2020 Annual

Prominence-based licensing in head movement and phrasal movement Brian Hsu LSA 2020 Annual Meeting January 4, 2020 1 Introduction and overview Focus of this talk: Some challenging interactions between head movement and phrasal movement.

557 views • 40 slides
1 SVD applications: rank, column, row, and null spaces Rank : the rank of a matrix is equal to:

1 SVD applications: rank, column, row, and null spaces Rank : the rank of a matrix is equal to:

Statistical Modeling and Analysis of Neural Data (NEU 560) Princeton University, Spring 2018 Jonathan Pillow Lecture 3A notes: SVD and Linear Systems 1 SVD applications: rank, column, row, and null spaces Rank : the rank of a matrix is equal to:

462 views • 3 slides
The The Sel Selecti ection n & Pres esentati entation n of Phrasal Verbs bs in ELT

The The Sel Selecti ection n & Pres esentati entation n of Phrasal Verbs bs in ELT

The The Sel Selecti ection n & Pres esentati entation n of Phrasal Verbs bs in ELT Textbo books: Alternative Appr pproaches Katie Heil March 7, 2015 NYS TESOL Applied Linguistics Winter Conference Teachers College, Columbia

869 views • 20 slides
Dependency and Phrasal Parsers of the Czech Language: A Comparison ak 1 , Tom s Holan 2 ,

Dependency and Phrasal Parsers of the Czech Language: A Comparison ak 1 , Tom s Holan 2 ,

Outline Motivation Compared Parsers Differences of the Parses Results Conclusions and Future Directions Dependency and Phrasal Parsers of the Czech Language: A Comparison ak 1 , Tom s Holan 2 , Vladim r Kadlec 1 , Vojt r 1 Ale s

734 views • 46 slides
WEBee Reverse Convolution Coding Reverse Convolution Coding Convolutional encoding uses a

WEBee Reverse Convolution Coding Reverse Convolution Coding Convolutional encoding uses a

WEBee Reverse Convolution Coding Reverse Convolution Coding Convolutional encoding uses a 288-by-216 matrix M M is not full row-rank (row:288 > column:216), the matrix equation is an overdetermined system ZigBee signals occupy

164 views • 15 slides
Chapter 2: Lexical and Phrasal Signs Syntactic Constructions in English Kim and Michaelis (2020)

Chapter 2: Lexical and Phrasal Signs Syntactic Constructions in English Kim and Michaelis (2020)

Chapter 2: Lexical and Phrasal Signs Syntactic Constructions in English Kim and Michaelis (2020) Syntactic Constructions Chapter 2 1 / 70 1 Linguistic Signs and Constructions as Form-Meaning Pairs 2 From Lexical Signs to Phrasal Signs as a

884 views • 70 slides
/k Content 2/15 1. Introduction 2. Hamming weight 3. Rank weight 4. Extended rank weight

/k Content 2/15 1. Introduction 2. Hamming weight 3. Rank weight 4. Extended rank weight

On defining the generalized rank weight Ruud Pellikaan joint work with Relinde Jurrius Autonomous University Barcelona, 6 November 2014 /k Content 2/15 1. Introduction 2. Hamming weight 3. Rank weight 4. Extended rank weight enumerator

165 views • 15 slides
Phrasal Restrictions on Noncontrastive Topic: The Case of Japanese Ross Kirsling Department of

Phrasal Restrictions on Noncontrastive Topic: The Case of Japanese Ross Kirsling Department of

Introduction Background Data Analysis Conclusion References Phrasal Restrictions on Noncontrastive Topic: The Case of Japanese Ross Kirsling Department of Linguistics University of Wisconsin-Madison Prelim 1 Defense December 6, 2012

1.06k views • 19 slides
10. Learning to Rank Outline 10.1. Why Learning to Rank (LeToR)? 10.2. Pointwise, Pairwise,

10. Learning to Rank Outline 10.1. Why Learning to Rank (LeToR)? 10.2. Pointwise, Pairwise,

10. Learning to Rank Outline 10.1. Why Learning to Rank (LeToR)? 10.2. Pointwise, Pairwise, Listwise 10.3. Gathering User Input 10.4. LeToR Evaluation 10.5. Beyond Search Advanced Topics in Information Retrieval / Learning to Rank 2 10.1.

1.01k views • 34 slides
Deep Encode: Machine Learning for Per-Title Encoding Daniel Silhavy| IBC20| Per-Title Encoding

Deep Encode: Machine Learning for Per-Title Encoding Daniel Silhavy| IBC20| Per-Title Encoding

Fraunhofer FOKUS Institut fr Offene Kommunikationssysteme Deep Encode: Machine Learning for Per-Title Encoding Daniel Silhavy| IBC20| Per-Title Encoding What & Why & 1 How? Benefits of Per-Title Encoding Storage and bitrate

342 views • 12 slides
Harmonic Grammar in phrasal movement: an account of probe competition and blocking Brian Hsu 51

Harmonic Grammar in phrasal movement: an account of probe competition and blocking Brian Hsu 51

Harmonic Grammar in phrasal movement: an account of probe competition and blocking Brian Hsu 51 st Annual Meeting of the North East Linguistic society (NELS 51) November 8, 2020 1 Organization of the talk 1. Overview 2. Probe competition and

714 views • 60 slides
61A Extra Lecture 4 Announcements Encoding Strings Representing Strings: UTF-8 Encoding 4

61A Extra Lecture 4 Announcements Encoding Strings Representing Strings: UTF-8 Encoding 4

61A Extra Lecture 4 Announcements Encoding Strings Representing Strings: UTF-8 Encoding 4 Representing Strings: UTF-8 Encoding UTF (UCS (Universal Character Set) Transformation Format) 4 Representing Strings: UTF-8 Encoding UTF (UCS

1.62k views • 99 slides
A new family of maximum rank distance codes or: Maximum rank distance codes and finite semifields

A new family of maximum rank distance codes or: Maximum rank distance codes and finite semifields

A new family of maximum rank distance codes or: Maximum rank distance codes and finite semifields John Sheekey Universiteit Gent, Belgium Alcoma, March 2015 Rank metric codes A rank metric code is a set C M n ( F ) of n n matrices over

963 views • 92 slides
Introduction to rank-based cryptography Philippe Gaborit University of Limoges, France ASCRYPTO

Introduction to rank-based cryptography Philippe Gaborit University of Limoges, France ASCRYPTO

Rank codes : definitions and basic properties Decoding in rank metric Complexity issues : decoding random rank codes Encryption Authentication and signature Introduction to rank-based cryptography Philippe Gaborit University of Limoges,

1.56k views • 115 slides
On the minimum rank of a graph Jisu Jeong June 21, 2013 Jisu Jeong On the minimum rank of a

On the minimum rank of a graph Jisu Jeong June 21, 2013 Jisu Jeong On the minimum rank of a

Minimum rank The minimum rank of a random graph over the binary field An algorithm to decide the minimum rank for fixed k Future work On the minimum rank of a graph Jisu Jeong June 21, 2013 Jisu Jeong On the minimum rank of a graph Minimum

865 views • 27 slides
Topic 9: Phrasal Verbs. Adolescence Adolescence Adolescence is the period of life between the

Topic 9: Phrasal Verbs. Adolescence Adolescence Adolescence is the period of life between the

Topic 9: Phrasal Verbs. Adolescence Adolescence Adolescence is the period of life between the beginning of puberty and adulthood. You previously learned about the physical changes of puberty. Adolescence is also a time of

149 views • 11 slides
Three Air Three Airpor ports Rank at ts Rank at the Bottom the Bottom At 20 At 20 minutes FAA

Three Air Three Airpor ports Rank at ts Rank at the Bottom the Bottom At 20 At 20 minutes FAA

Report Findings Three Air Three Airpor ports Rank at ts Rank at the Bottom the Bottom At 20 At 20 minutes FAA caps flights, which limits use At 10 At 10 minutes to match most major airports, improving level of service Air Passengers

809 views • 47 slides
A Compositional Approach toward Dynamic Phrasal Thesaurus Atsushi FUJITA, Shuhei KATO, Naoki

A Compositional Approach toward Dynamic Phrasal Thesaurus Atsushi FUJITA, Shuhei KATO, Naoki

< WTEP 2007, Jun. 29th, 2007 > A Compositional Approach toward Dynamic Phrasal Thesaurus Atsushi FUJITA, Shuhei KATO, Naoki KATO, Satoshi SATO Nagoya Univ., Japan Computing Semantic Equivalence (SE) Fundamental in NLP Recognition:

630 views • 33 slides
Top Rank SEO Philippines was established in 2015 by Search Engine Optimization Specialists with

Top Rank SEO Philippines was established in 2015 by Search Engine Optimization Specialists with

Start Your Success With Us! Top Rank SEO Philippines is a group of highly skilled specialists who will help your website rank higher in search engines, allowing bigger ROI for you. Top Rank SEO Philippines was established in 2015 by Search

351 views • 5 slides
Instruction Encoding CSE378 W INTER , 2001 63 Introduction MIPS Encoding Remember that in a

Instruction Encoding CSE378 W INTER , 2001 63 Introduction MIPS Encoding Remember that in a

Instruction Encoding CSE378 W INTER , 2001 63 Introduction MIPS Encoding Remember that in a stored program computer, instructions are The nice thing about MIPS (and other RISC machines) is that it stored in memory (just like data) has

157 views • 4 slides

More recommend