Machine Translation Course 9 Diana Trandab ă ț Academic year 2014-2015.

Slides:



Advertisements
Similar presentations
Statistical Machine Translation
Advertisements

Solving problems by searching
Chapter 5: Tree Constructions
Heuristic Search techniques
Announcements Course TA: Danny Kumar
Conceptual Clustering
Statistical Machine Translation Part II: Word Alignments and EM Alexander Fraser Institute for Natural Language Processing University of Stuttgart
Statistical Machine Translation Part II: Word Alignments and EM Alexander Fraser ICL, U. Heidelberg CIS, LMU München Statistical Machine Translation.
Statistical Machine Translation Part II – Word Alignments and EM Alex Fraser Institute for Natural Language Processing University of Stuttgart
1 EE5900 Advanced Embedded System For Smart Infrastructure Static Scheduling.
Techniques for Dealing with Hard Problems Backtrack: –Systematically enumerates all potential solutions by continually trying to extend a partial solution.
Problem Solving and Search Andrea Danyluk September 9, 2013.
Graphs Graphs are the most general data structures we will study in this course. A graph is a more general version of connected nodes than the tree. Both.
Best-First Search: Agendas
Module 1 Dictionary skills Part 1
Search: Representation and General Search Procedure Jim Little UBC CS 322 – Search 1 September 10, 2014 Textbook § 3.0 –
Hidden Markov Models Bonnie Dorr Christof Monz CMSC 723: Introduction to Computational Linguistics Lecture 5 October 6, 2004.
3 -1 Chapter 3 The Greedy Method 3 -2 The greedy method Suppose that a problem can be solved by a sequence of decisions. The greedy method has that each.
Using Search in Problem Solving
CSE 780 Algorithms Advanced Algorithms Graph Algorithms Representations BFS.
Using Search in Problem Solving
מבנה המחשב – מבוא למחשבים ספרתיים Foundations of Combinational Circuits תרגול מספר 3.
Backtracking Reading Material: Chapter 13, Sections 1, 2, 4, and 5.
Data Flow Analysis Compiler Design Nov. 8, 2005.
Solving problems by searching
© 2005 by Prentice Hall Chapter 9 Structuring System Requirements: Logic Modeling Modern Systems Analysis and Design Fourth Edition.
1 Statistical NLP: Lecture 13 Statistical Alignment and Machine Translation.
1 Shortest Path Calculations in Graphs Prof. S. M. Lee Department of Computer Science.
Natural Language Processing Lab Northeastern University, China Feiliang Ren EBMT Based on Finite Automata State Transfer Generation Feiliang Ren.
Machine translation Context-based approach Lucia Otoyo.
English-Persian SMT Reza Saeedi 1 WTLAB Wednesday, May 25, 2011.
Artificial Intelligence Introduction (2). What is Artificial Intelligence ?  making computers that think?  the automation of activities we associate.
© The McGraw-Hill Companies, Inc., Chapter 3 The Greedy Method.
Statistical Machine Translation Part IV – Log-Linear Models Alex Fraser Institute for Natural Language Processing University of Stuttgart Seminar:
Advanced Signal Processing 05/06 Reinisch Bernhard Statistical Machine Translation Phrase Based Model.
Ch.12 Machine Learning Genetic Algorithm Dr. Bernard Chen Ph.D. University of Central Arkansas Spring 2011.
Introduction to search Chapter 3. Why study search? §Search is a basis for all AI l search proposed as the basis of intelligence l inference l all learning.
Introduction to search Chapter 3. Why study search? §Search is a basis for all AI l search proposed as the basis of intelligence l all learning algorithms,
Dr.Abeer Mahmoud ARTIFICIAL INTELLIGENCE (CS 461D) Dr. Abeer Mahmoud Computer science Department Princess Nora University Faculty of Computer & Information.
CS 415 – A.I. Slide Set 5. Chapter 3 Structures and Strategies for State Space Search – Predicate Calculus: provides a means of describing objects and.
Flowcharts. Problem Solving Computer programs are written to solve problems or perform tasks Programmers translate the solutions or tasks into a language.
Statistical Machine Translation Part III – Phrase-based SMT / Decoding Alexander Fraser Institute for Natural Language Processing Universität Stuttgart.
1 Solving problems by searching 171, Class 2 Chapter 3.
Search CPSC 386 Artificial Intelligence Ellen Walker Hiram College.
Lecture 3: 18/4/1435 Searching for solutions. Lecturer/ Kawther Abas 363CS – Artificial Intelligence.
AI Lecture 17 Planning Noémie Elhadad (substituting for Prof. McKeown)
Haitham Elmarakeby.  Speech recognition
1 Minimum Error Rate Training in Statistical Machine Translation Franz Josef Och Information Sciences Institute University of Southern California ACL 2003.
Introduction to Artificial Intelligence (G51IAI) Dr Rong Qu Blind Searches - Introduction.
Chapter 12 search and speaker adaptation 12.1 General Search Algorithm 12.2 Search Algorithms for Speech Recognition 12.3 Language Model States 12.4 Speaker.
Statistical Machine Translation Part II: Word Alignments and EM Alex Fraser Institute for Natural Language Processing University of Stuttgart
COSC 3101A - Design and Analysis of Algorithms 14 NP-Completeness.
Spring 2010 Lecture 4 Kristina Toutanova MSR & UW With slides borrowed from Philipp Koehn and Hwee Tou Ng LING 575: Seminar on statistical machine translation.
Solving problems by searching Chapter 3. Types of agents Reflex agent Consider how the world IS Choose action based on current percept Do not consider.
Neural Machine Translation
Statistical Machine Translation Part II: Word Alignments and EM
Solving problems by searching
ECE 448 Lecture 4: Search Intro
Statistical NLP: Lecture 13
Ordering of Hypothesis Space
Neural Lattice Search for Domain Adaptation in Machine Translation
CSCI 5832 Natural Language Processing
Statistical Machine Translation Part III – Phrase-based SMT / Decoding
Chapter 9 Structuring System Requirements: Logic Modeling
CIS 488/588 Bruce R. Maxim UM-Dearborn
Statistical Machine Translation Part IIIb – Phrase-based Model
Chapter 9 Structuring System Requirements: Logic Modeling
Statistical Machine Translation Part VI – Phrase-based Decoding
Implementation of Learning Systems
Neural Machine Translation by Jointly Learning to Align and Translate
Presentation transcript:

Machine Translation Course 9 Diana Trandab ă ț Academic year

Statistical Machine Translation Bilingual text Monolingual text Translation model P(s|t) Language model P(t) Decoding algorithm argmax P(t)*P(s|t)

Decoding How to translate new sentences? Probability models enable us to make predictions: – Given a particular English sentence, what is the most probable Foreign sentence corresponding to it? In math, we want to solve: argmax English p(English|Foreign) Problem: there are a lot of Foreign possible sentences to choose from!!! In combination with a language model, the decoder generates the most likely translation.

Little story... A French speaking traveler equipped with a bilingual dictionary enters in a New-York store. While reviewing the price chart, he encounters a line that he does not understand: A sheet of paper $ We will look at a process this traveler could use to decode this strange sentence.

Resources Sentence to translate: A sheet of paper Bilingual dictionary The traveler’s common sense: The traveller can intuitively evaluate the likelihood of a French sentence (since he is French). sourcetargetsourcetarget AUnofde AUneofdu sheetfeuillepaperpapier sheetdrap

The traveler’s decoder With these resources in hand, the traveler can use the following algorithm to translate the sentence one word at a time: initialize the set of candidate translations H with an empty sentence. while there are incomplete sentences in H – pick the least completed translation h from H – for each possible translation for the next word to translate in h append at the end of h and store the result in hcopy if hcopy is likely following the traveler’s intuition, add it to H return the best candidate in H

Search graph The traveler concludes that the most likely translation is Une feuille de papier.

Search space complexity

Decoder A decoder searches the target document t having the highest probability to translate a given source document s. Most decoders assume that the sentences of a document are independents one from each others. The decoder can thus translate each sentence individually. Shortcomings: – A sentence cannot be omitted, merged with another one, repositioned or sliced by the decoder. – The context of a sentence is not considered when it is translated.

The decoder’s task The task of the decoder is to use its knowledge and a density function to find the best sequence of transformations that can be applied to an initial target sentence to translate a given source sentence. This problem can be reformulated as a classic AI problem: searching for the shortest path in an implicit graph. Two independent problems must be resolved in order to build a decoder: – Model representation The model defines what a transformation is and how to evaluate the quality of a translation. – Search space exploration Enumerating all possible sequences of transformations is often impracticable, we must smartly select the ones that will be evaluated.

Model representation — Partial translation The partial translation is a translation that is being transformed. It is composed of: – the source sentence – the target sentence that is being built – a progress indicator that shows how to continue this translation The source and target sentences can be sequences of words, trees, non-contiguous sentences, etc.

Model representation — Transformation A partial translation evolves when a transformation is applied to it. A transformation can take many forms: – translation of one word – translation of many words – reordering of the children of a node in a tree

Model representation — Cost A cost quantify the quality of a partial translation. Usually, it evaluates at least: – the likelihood of the transformations – the fluency of the target sentence generated so far – the word reordering that occurred The cost is used to identify the partial translations to dismiss and to select the best complete translation when the search ends.

Model representation — Transformation generator The transformation generator takes as input a partial translation and outputs the set of transformations that can be applied to it.

Search space exploration — Hypothesis and search strategy A hypothesis is made of a partial translation and of a cost. The task of the search strategy is two-fold: – Deciding the order in which the hypotheses are explored. – Identify the hypotheses to dismiss (using the value of the cost).

Putting it all together Each vertex is a hypothesis (partial translation and cost) Each edge corresponds to a transformation. The transformation generator enumerates the out-edges of each vertex. The search strategy defines how to explore the graph

Slide from Koehn 2008

Modified from Koehn 2008 (possible future paths are the same)

Modified from Koehn 2008 (possible future paths are the same)

Slide from Koehn 2008

Great! See you next time!