Skip to content
项目
群组
代码片段
帮助
当前项目
正在载入...
登录 / 注册
切换导航面板
T
Toy-MT-Introduction
概览
Overview
Details
Activity
Cycle Analytics
版本库
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
问题
0
Issues
0
列表
Board
标记
里程碑
合并请求
0
Merge Requests
0
CI / CD
CI / CD
流水线
作业
日程表
图表
维基
Wiki
代码片段
Snippets
成员
Collapse sidebar
Close sidebar
活动
图像
聊天
创建新问题
作业
提交
Issue Boards
Open sidebar
单韦乔
Toy-MT-Introduction
Commits
1e144093
Commit
1e144093
authored
Nov 18, 2019
by
xiaotong
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
new pages and new sections
parent
65c39211
显示空白字符变更
内嵌
并排
正在显示
5 个修改的文件
包含
138 行增加
和
696 行删除
+138
-696
.gitignore
+1
-0
Section04-Phrasal-and-Syntactic-Models/section04.test.tex
+0
-442
Section04-Phrasal-and-Syntactic-Models/section04.tex
+95
-93
Section06-Neural-Machine-Translation/section06-test.tex
+18
-161
Section06-Neural-Machine-Translation/section06.tex
+24
-0
没有找到文件。
.gitignore
查看文件 @
1e144093
...
...
@@ -7,3 +7,4 @@
*.gz
*.toc
*.blg
*.sav
Section04-Phrasal-and-Syntactic-Models/section04.test.tex
deleted
100644 → 0
查看文件 @
65c39211
% !Mode:: "TeX:GBK"
\def\CTeXPreproc
{
Created by ctex v0.2.13, don't edit!
}
\documentclass
[cjk,t,compress,12pt]
{
beamer
}
%\documentclass{article}
%\usepackage{beamerarticle}
\usepackage
{
pstricks
}
\usepackage
{
etex
}
\usepackage
{
eso-pic,graphicx
}
\usepackage
{
fancybox
}
\usepackage
{
amsmath,amssymb
}
\usepackage
{
setspace
}
\usepackage
{
xcolor
}
\usepackage
{
CJK
}
\usepackage
{
tikz
}
\usepackage
{
tikz-qtree
}
\usepackage
{
hyperref
}
\usepackage
{
ulem
}
\usetikzlibrary
{
arrows,decorations.pathreplacing
}
\usetikzlibrary
{
shadows
}
% LATEX and plain TEX when using Tik Z
\usepgflibrary
{
arrows
}
% LATEX and plain TEX and pure pgf
\usetikzlibrary
{
arrows
}
% LATEX and plain TEX when using Tik Z
\usetikzlibrary
{
decorations
}
\usetikzlibrary
{
arrows,shapes
}
\usetikzlibrary
{
decorations.text
}
\usetikzlibrary
{
positioning,fit,calc
}
\usetikzlibrary
{
mindmap,backgrounds
}
% mind map
\DeclareMathOperator*
{
\argmax
}{
arg
\,
max
}
\DeclareMathOperator*
{
\argmin
}{
arg
\,
min
}
\setbeamertemplate
{
items
}
[ball]
\usefonttheme
[onlymath]
{
serif
}
% fout of math
\definecolor
{
ugreen
}{
rgb
}{
0,0.5,0
}
\definecolor
{
lgreen
}{
rgb
}{
0.9,1,0.8
}
\definecolor
{
xtgreen1
}{
rgb
}{
0.824,0.898,0.8
}
\definecolor
{
xtgreen
}{
rgb
}{
0.914,0.945,0.902
}
\definecolor
{
lightgray
}{
gray
}{
0.85
}
\setbeamercolor
{
uppercol
}{
fg=white,bg=ugreen
}
\setbeamercolor
{
lowercol
}{
fg=black,bg=xtgreen
}
%\definecolor{ublue}{rgb}{0,0.298,0.525}
\definecolor
{
ublue
}{
rgb
}{
0.152,0.250,0.545
}
\setbeamercolor
{
uppercolblue
}{
fg=white,bg=ublue
}
\setbeamercolor
{
lowercolblue
}{
fg=black,bg=blue!10
}
%\usetheme{default}
%\usetheme{Darmstadt}
%\usetheme{Madrid}
%\usetheme{Frankfurt}
%\usetheme{Dresden}
%\usetheme{Boadilla}
%\usecolortheme{dolphin}
\usefonttheme
[onlylarge]
{
structurebold
}
\begin{CJK}
{
GBK
}{
song
}
\end{CJK}
\setbeamerfont*
{
frametitle
}{
size=
\large
,series=
\bfseries
}
\setbeamertemplate
{
navigation symbols
}{
\begin{CJK}
{
GBK
}{
hei
}
第四章 基于短语的翻译模型
\hspace*
{
2em
}
肖桐
\&
朱靖波
\end{CJK}
\hspace*
{
2em
}
\today
\hspace*
{
2em
}
\insertframenumber
{}
/
\inserttotalframenumber
}
\setbeamertemplate
{
itemize items
}
[circle]
% if you want a circle
\setbeamertemplate
{
itemize subitem
}
[triangle]
% if you want a triangle
\setbeamertemplate
{
itemize subsubitem
}
[ball]
% if you want a ball
\begin{document}
\begin{CJK}
{
GBK
}{
you
}
\title
{
\Large
{
基于短语的翻译模型
}}
\author
{
\large
{
\textbf
{
肖桐
\ \
朱靖波
}}}
\institute
{
\blue
{
\url
{
xiaotong@mail.neu.edu.cn
}}
\black
{}
\\
\blue
{
\url
{
zhujingbo@mail.neu.edu.cn
}}
\black
{}
\\
\vspace
{
1.0em
}
东北大学 自然语言处理实验室
\\
\blue
{
\underline
{
\url
{
http://www.nlplab.com
}}}
\black
{}
\\
\vspace
{
0.2cm
}
\hspace
{
0.1cm
}
\includegraphics
[scale=0.1]
{
../Figures/logo.pdf
}
}
\date
{}
\maketitle
\setlength
{
\leftmargini
}{
1em
}
\setlength
{
\leftmarginii
}{
1em
}
%%%------------------------------------------------------------------------------------------------------------
\section
{
翻译中为什么要使用短语
}
%%%------------------------------------------------------------------------------------------------------------
%%% 回顾基于词的翻译模型
\begin{frame}
{
回顾 - 基于词的机器翻译
}
\begin{itemize}
\item
对每个
\alert
{
单词
}
的翻译进行(任意)组合
\end{itemize}
\vspace
{
-2em
}
\begin{flushright}
\begin{tikzpicture}
\begin{scope}
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词):
}}}}
;
\draw
[->,very thick,ublue] (s1.south) -- ([yshift=-0.7em]s1.south);
\draw
[->,very thick,ublue] (s2.south) -- ([yshift=-0.7em]s2.south);
\draw
[->,very thick,ublue] (s3.south) -- ([yshift=-0.7em]s3.south);
\draw
[->,very thick,ublue] (s4.south) -- ([yshift=-0.7em]s4.south);
\draw
[->,very thick,ublue] (s5.south) -- ([yshift=-0.7em]s5.south);
{
\small
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t11) at ([yshift=-1em]s1.south)
{
I
}
;
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t12) at ([yshift=-0.2em]t11.south)
{
me
}
;
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t13) at ([yshift=-0.2em]t12.south)
{
I'm
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl11) at (t11.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl12) at (t12.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl13) at (t13.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\invisible
{
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=6.55em] (t14) at ([yshift=-0.2em]t13.south west)
{
I'm
}
;
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=6.55em] (t15) at ([yshift=-0.2em]t14.south west)
{
I
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl14) at (t14.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1-2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl15) at (t15.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1-2
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t21) at ([yshift=-1em]s2.south)
{
to
}
;
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t22) at ([yshift=-0.2em]t21.south)
{
with
}
;
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t23) at ([yshift=-0.2em]t22.south)
{
for
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl21) at (t21.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl22) at (t22.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl23) at (t23.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\invisible
{
\node
[anchor=north west,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=6.55em] (t24) at ([yshift=-0.2em,xshift=-2.6em]t15.south east)
{
for you
}
;
\node
[anchor=north west,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=6.55em] (t25) at ([yshift=-0.2em]t24.south west)
{
with you
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl24) at (t24.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2-3
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl25) at (t25.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2-3
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=2.5em] (t31) at ([yshift=-1em]s3.south)
{
you
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl31) at (t31.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3
}}}}
;
\invisible
{
\node
[anchor=west,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=13.35em] (t32) at ([xshift=1.4em]t14.east)
{
you are satisfied
}
;
\node
[anchor=north west,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=7.45em] (t33) at ([yshift=-0.2em]t32.south west)
{$
\phi
$}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl32) at (t32.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3-5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl33) at (t33.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3-4
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=orange!20,minimum height=1.5em,minimum width=3em] (t41) at ([yshift=-1em]s4.south)
{$
\phi
$}
;
\node
[anchor=north,inner sep=2pt,fill=orange!20,minimum height=1.5em,minimum width=3em] (t42) at ([yshift=-0.2em]t41.south)
{
show
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl41) at (t41.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl42) at (t42.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4
}}}}
;
\invisible
{
\node
[anchor=west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=9.00em] (t43) at ([xshift=1.75em]t24.east)
{
satisfied
}
;
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=9.00em] (t44) at ([yshift=-0.2em]t43.south west)
{
satisfactory
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl43) at (t43.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4-5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl44) at (t44.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4-5
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t51) at ([yshift=-1em]s5.south)
{
satisfy
}
;
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t52) at ([yshift=-0.2em]t51.south)
{
satisfied
}
;
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t53) at ([yshift=-0.2em]t52.south)
{
satisfies
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl51) at (t51.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl52) at (t52.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl53) at (t53.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
}
{
\tiny
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt11) at (t11.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt12) at (t12.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt13) at (t13.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\invisible
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt14) at (t14.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt15) at (t15.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt21) at (t21.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt22) at (t22.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt23) at (t23.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\invisible
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt24) at (t24.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt25) at (t25.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt31) at (t31.east)
{{
\color
{
white
}
\textbf
{
P=1
}}}
;
\invisible
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt33) at (t32.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt33) at (t33.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt41) at (t41.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt42) at (t42.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\invisible
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt43) at (t43.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt44) at (t44.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt51) at (t51.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt52) at (t52.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt53) at (t53.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
}
\end{scope}
\begin{scope}
{
\small
}
\end{scope}
\begin{scope}
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}}
;
\visible
<2->
{
\draw
[->,ultra thick,red,line width=2pt,opacity=0.7] ([xshift=-0.5em]t13.west) -- ([xshift=0.8em]t13.east) -- ([xshift=-0.2em]t22.west) -- ([xshift=0.8em]t22.east) -- ([xshift=-0.2em]t31.west) -- ([xshift=0.8em]t31.east) -- ([xshift=-0.2em]t41.west) -- ([xshift=0.8em]t41.east) -- ([xshift=-0.2em]t52.west) -- ([xshift=1.2em]t52.east);
}
\invisible
{
\draw
[->,ultra thick,ublue,line width=2pt,opacity=0.7] ([xshift=-0.5em]t15.west) -- ([xshift=0.8em]t15.east) -- ([xshift=-0.2em]t32.west) -- ([xshift=1.2em]t32.east);
\draw
[->,ultra thick,ublue,line width=2pt,opacity=0.7] ([xshift=-0.5em,yshift=0.1em]t13.west) -- ([xshift=0.8em,yshift=0.1em]t13.east) -- ([xshift=-0.2em]t25.west) -- ([xshift=0.8em]t25.east) -- ([xshift=-0.2em,yshift=0.1em]t41.west) -- ([xshift=0.8em,yshift=0.1em]t41.east) -- ([xshift=-0.2em,yshift=0.1em]t52.west) -- ([xshift=1.2em,yshift=0.1em]t52.east);
}
\invisible
{
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}}
;
}
\visible
<2->
{
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词):
}}
;
\draw
[->,ultra thick,red,line width=1.5pt,opacity=0.7] (wtranslabel.east) -- ([xshift=1em]wtranslabel.east);
}
\invisible
{
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语):
}}
;
\draw
[->,ultra thick,ublue,line width=1.5pt,opacity=0.7] ([xshift=0.65em]ptranslabel.east) -- ([xshift=1.65em]ptranslabel.east);
}
\end{scope}
\end{tikzpicture}
\end{flushright}
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
%%% 基于短语的翻译模型
\begin{frame}
{
基于"短语"的机器翻译
}
\begin{itemize}
\item
对每个
\alert
{
单词
}
及
\alert
{
连续词串
}
的翻译进行(任意)组合
\end{itemize}
\vspace
{
-2em
}
\begin{flushright}
\begin{tikzpicture}
\begin{scope}
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词):
}}}}
;
\draw
[->,very thick,ublue] (s1.south) -- ([yshift=-0.7em]s1.south);
\draw
[->,very thick,ublue] (s2.south) -- ([yshift=-0.7em]s2.south);
\draw
[->,very thick,ublue] (s3.south) -- ([yshift=-0.7em]s3.south);
\draw
[->,very thick,ublue] (s4.south) -- ([yshift=-0.7em]s4.south);
\draw
[->,very thick,ublue] (s5.south) -- ([yshift=-0.7em]s5.south);
{
\small
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t11) at ([yshift=-1em]s1.south)
{
I
}
;
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t12) at ([yshift=-0.2em]t11.south)
{
me
}
;
\node
[anchor=north,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=2.5em] (t13) at ([yshift=-0.2em]t12.south)
{
I'm
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl11) at (t11.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl12) at (t12.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl13) at (t13.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1
}}}}
;
\visible
<2->
{
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=6.55em] (t14) at ([yshift=-0.2em]t13.south west)
{
I'm
}
;
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=6.55em] (t15) at ([yshift=-0.2em]t14.south west)
{
I
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl14) at (t14.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1-2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl15) at (t15.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
1-2
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t21) at ([yshift=-1em]s2.south)
{
to
}
;
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t22) at ([yshift=-0.2em]t21.south)
{
with
}
;
\node
[anchor=north,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=2.5em] (t23) at ([yshift=-0.2em]t22.south)
{
for
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl21) at (t21.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl22) at (t22.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl23) at (t23.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2
}}}}
;
\visible
<2->
{
\node
[anchor=north west,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=6.55em] (t24) at ([yshift=-0.2em,xshift=-2.6em]t15.south east)
{
for you
}
;
\node
[anchor=north west,inner sep=2pt,fill=green!20,minimum height=1.5em,minimum width=6.55em] (t25) at ([yshift=-0.2em]t24.south west)
{
with you
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl24) at (t24.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2-3
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl25) at (t25.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
2-3
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=2.5em] (t31) at ([yshift=-1em]s3.south)
{
you
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl31) at (t31.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3
}}}}
;
\visible
<2->
{
\node
[anchor=west,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=13.35em] (t32) at ([xshift=1.4em]t14.east)
{
you are satisfied
}
;
\node
[anchor=north west,inner sep=2pt,fill=blue!20,minimum height=1.5em,minimum width=7.45em] (t33) at ([yshift=-0.2em]t32.south west)
{$
\phi
$}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl32) at (t32.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3-5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl33) at (t33.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
3-4
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=orange!20,minimum height=1.5em,minimum width=3em] (t41) at ([yshift=-1em]s4.south)
{$
\phi
$}
;
\node
[anchor=north,inner sep=2pt,fill=orange!20,minimum height=1.5em,minimum width=3em] (t42) at ([yshift=-0.2em]t41.south)
{
show
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl41) at (t41.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl42) at (t42.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4
}}}}
;
\visible
<2->
{
\node
[anchor=west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=9.00em] (t43) at ([xshift=1.75em]t24.east)
{
satisfied
}
;
\node
[anchor=north west,inner sep=2pt,fill=red!20,minimum height=1.5em,minimum width=9.00em] (t44) at ([yshift=-0.2em]t43.south west)
{
satisfactory
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl43) at (t43.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4-5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl44) at (t44.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
4-5
}}}}
;
}
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t51) at ([yshift=-1em]s5.south)
{
satisfy
}
;
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t52) at ([yshift=-0.2em]t51.south)
{
satisfied
}
;
\node
[anchor=north,inner sep=2pt,fill=purple!20,minimum height=1.5em,minimum width=4.5em] (t53) at ([yshift=-0.2em]t52.south)
{
satisfies
}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl51) at (t51.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl52) at (t52.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
\node
[anchor=north west,inner sep=1pt,fill=black] (tl53) at (t53.north west)
{
\tiny
{{
\color
{
white
}
\textbf
{
5
}}}}
;
}
{
\tiny
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt11) at (t11.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt12) at (t12.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt13) at (t13.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\visible
<2->
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt14) at (t14.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt15) at (t15.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt21) at (t21.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt22) at (t22.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt23) at (t23.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\visible
<2->
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt24) at (t24.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt25) at (t25.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt31) at (t31.east)
{{
\color
{
white
}
\textbf
{
P=1
}}}
;
\visible
<2->
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt33) at (t32.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt33) at (t33.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt41) at (t41.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt42) at (t42.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\visible
<2->
{
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt43) at (t43.east)
{{
\color
{
white
}
\textbf
{
P=.3
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt44) at (t44.east)
{{
\color
{
white
}
\textbf
{
P=.2
}}}
;
}
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt51) at (t51.east)
{{
\color
{
white
}
\textbf
{
P=.5
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt52) at (t52.east)
{{
\color
{
white
}
\textbf
{
P=.4
}}}
;
\node
[anchor=north,rotate=90,inner sep=1pt,minimum width=2.55em,fill=black] (pt53) at (t53.east)
{{
\color
{
white
}
\textbf
{
P=.1
}}}
;
}
\end{scope}
\begin{scope}
{
\small
}
\end{scope}
\begin{scope}
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}}
;
\draw
[->,ultra thick,red,line width=2pt,opacity=0.7] ([xshift=-0.5em]t13.west) -- ([xshift=0.8em]t13.east) -- ([xshift=-0.2em]t22.west) -- ([xshift=0.8em]t22.east) -- ([xshift=-0.2em]t31.west) -- ([xshift=0.8em]t31.east) -- ([xshift=-0.2em]t41.west) -- ([xshift=0.8em]t41.east) -- ([xshift=-0.2em]t52.west) -- ([xshift=1.2em]t52.east);
\visible
<3->
{
\draw
[->,ultra thick,ublue,line width=2pt,opacity=0.7] ([xshift=-0.5em]t15.west) -- ([xshift=0.8em]t15.east) -- ([xshift=-0.2em]t32.west) -- ([xshift=1.2em]t32.east);
}
\visible
<4->
{
\draw
[->,ultra thick,ublue,line width=2pt,opacity=0.7] ([xshift=-0.5em,yshift=0.1em]t13.west) -- ([xshift=0.8em,yshift=0.1em]t13.east) -- ([xshift=-0.2em]t25.west) -- ([xshift=0.8em]t25.east) -- ([xshift=-0.2em,yshift=0.1em]t41.west) -- ([xshift=0.8em,yshift=0.1em]t41.east) -- ([xshift=-0.2em,yshift=0.1em]t52.west) -- ([xshift=1.2em,yshift=0.1em]t52.east);
}
\visible
<2->
{
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}}
;
}
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词):
}}
;
\draw
[->,ultra thick,red,line width=1.5pt,opacity=0.7] (wtranslabel.east) -- ([xshift=1em]wtranslabel.east);
\visible
<3->
{
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语):
}}
;
\draw
[->,ultra thick,ublue,line width=1.5pt,opacity=0.7] ([xshift=0.65em]ptranslabel.east) -- ([xshift=1.65em]ptranslabel.east);
}
\end{scope}
\end{tikzpicture}
\end{flushright}
\end{frame}
\end{CJK}
\end{document}
Section04-Phrasal-and-Syntactic-Models/section04.tex
查看文件 @
1e144093
% !Mode:: "TeX:GBK"
% !Mode:: "TeX:UTF-8"
% !TEX encoding = UTF-8 Unicode
\def\CTeXPreproc
{
Created by ctex v0.2.13, don't edit!
}
\documentclass
[cjk,t,compress,12pt]
{
beamer
}
%\documentclass[cjk,t,compress,handout,12pt]{beamer}
%\documentclass{article}
%\usepackage{beamerarticle}
\usepackage
{
pstricks
}
\usepackage
{
etex
}
\usepackage
{
eso-pic,graphicx
}
...
...
@@ -62,11 +60,15 @@
\usefonttheme
[onlylarge]
{
structurebold
}
\begin{CJK}
{
GBK
}{
song
}
\IfFileExists
{
C:/WINDOWS/win.ini
}
{
\newcommand
{
\mycfont
}{
you
}}
{
\newcommand
{
\mycfont
}{
gbsn
}}
\begin{CJK}
{
UTF8
}{
\mycfont
}
\end{CJK}
\setbeamerfont*
{
frametitle
}{
size=
\large
,series=
\bfseries
}
\setbeamertemplate
{
navigation symbols
}{
\begin{CJK}
{
GBK
}{
hei
}
第四章 基于短语和句法的翻译模型
\hspace*
{
2em
}
肖桐
\&
朱靖波
\end{CJK}
\hspace*
{
2em
}
\today
\hspace*
{
2em
}
\insertframenumber
{}
/
\inserttotalframenumber
}
\setbeamertemplate
{
navigation symbols
}{
\begin{CJK}
{
UTF8
}{
\mycfont
}
第四章 基于短语和句法的翻译模型
\hspace*
{
2em
}
肖桐
\&
朱靖波
\end{CJK
}
\hspace*
{
2em
}
\today
\hspace*
{
2em
}
\insertframenumber
{}
/
\inserttotalframenumber
}
\setbeamertemplate
{
itemize items
}
[circle]
% if you want a circle
\setbeamertemplate
{
itemize subitem
}
[triangle]
% if you wnat a triangle
...
...
@@ -74,15 +76,15 @@
\begin{document}
\begin{CJK}
{
GBK
}{
you
}
\begin{CJK}
{
UTF8
}{
\mycfont
}
\title
{
\Large
{
基于短语和句法的翻译模型
}}
\author
{
\large
{
\textbf
{
肖桐
\ \
朱靖波
}
}}
\title
{
\Large
{
基于短语和句法的翻译模型
}}
\author
{
\large
{
\textbf
{
肖桐
\ \
朱靖波
}}}
\institute
{
\blue
{
\url
{
xiaotong@mail.neu.edu.cn
}}
\black
{}
\\
\blue
{
\url
{
zhujingbo@mail.neu.edu.cn
}}
\black
{}
\\
\vspace
{
1.0em
}
东北大学 自然语言处理实验室
\\
东北大学 自然语言处理实验室
\\
\blue
{
\underline
{
\url
{
http://www.nlplab.com
}}}
\black
{}
\\
\vspace
{
0.2cm
}
\hspace
{
0.1cm
}
\includegraphics
[scale=0.1]
{
../Figures/logo.pdf
}
...
...
@@ -95,17 +97,17 @@
\setlength
{
\leftmarginii
}{
1em
}
%%%------------------------------------------------------------------------------------------------------------
\section
{
翻译中为什么要使用短语
}
\section
{
翻译中为什么要使用短语
}
%%%------------------------------------------------------------------------------------------------------------
%%%
基于单词的模型的问题
\begin{frame}
{
基于单词有哪些问题
}
%%%
基于单词的模型的问题
\begin{frame}
{
基于单词有哪些问题
}
\begin{itemize}
\item
基于单词的翻译模型有哪些
\alert
{
优点
}
?
\item
基于单词的翻译模型有哪些
\alert
{
优点
}
?
\begin{itemize}
\item
比较符合人的思维、简单直接、易于实现
\item
比较符合人的思维、简单直接、易于实现
\end{itemize}
\vspace
{
0.3em
}
...
...
@@ -114,11 +116,11 @@
\begin{center}
{
\scriptsize
\begin{tabular}
{
l | l
}
单词翻译表
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
绿
$
\to
$
green
&
0.9
\\
茶
$
\to
$
tea
&
0.8
\\
单词翻译表
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
绿
$
\to
$
green
&
0.9
\\
茶
$
\to
$
tea
&
0.8
\\
\end{tabular}
}
\end{center}
...
...
@@ -128,10 +130,10 @@
\begin{scope}
{
\footnotesize
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{{
\color
{
ugreen
}
绿
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{{
\color
{
ugreen
}
绿
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=east] (s) at (s1.west)
{$
s
=
$}
;
}
\end{scope}
...
...
@@ -158,11 +160,11 @@
\vspace
{
0.5em
}
\item
<3->
基于单词的翻译模型有哪些
\alert
{
缺点
}
?
\item
<3->
基于单词的翻译模型有哪些
\alert
{
缺点
}
?
\begin{itemize}
\item
很多 - 这符合中国人的思维,缺点总是很多
:)
\item
独立性假设:单词之间相对独立,没有考虑搭配
\item
调序:较弱的调序建模
\item
很多 - 这符合中国人的思维,缺点总是很多 :)
\item
独立性假设:单词之间相对独立,没有考虑搭配
\item
调序:较弱的调序建模
\item
...
\end{itemize}
...
...
@@ -172,12 +174,12 @@
\begin{center}
{
\scriptsize
\begin{tabular}
{
l | l
}
单词翻译表
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
红
$
\to
$
red
&
0.8
\\
红
$
\to
$
black
&
0.1
\\
茶
$
\to
$
tea
&
0.8
\\
单词翻译表
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
红
$
\to
$
red
&
0.8
\\
红
$
\to
$
black
&
0.1
\\
茶
$
\to
$
tea
&
0.8
\\
\end{tabular}
}
\end{center}
...
...
@@ -188,10 +190,10 @@
\begin{scope}
{
\footnotesize
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=east] (s) at (s1.west)
{$
s
=
$}
;
}
\end{scope}
...
...
@@ -216,7 +218,7 @@
\begin{pgfonlayer}
{
background
}
\visible
<5->
{
\node
[rectangle,draw=red,thick,inner sep=0.2em,fill=white,drop shadow] [fit = (t3) (t4)] (problemphrase)
{}
;
\node
[anchor=north,red,text width=8em,align=left] (problemlabel) at (problemphrase.south)
{
\begin{spacing}
{
0.8
}
\scriptsize
{
"
红 茶"为一种搭配,应该翻译为
"black tea"
}
\end{spacing}
}
;
\node
[anchor=north,red,text width=8em,align=left] (problemlabel) at (problemphrase.south)
{
\begin{spacing}
{
0.8
}
\scriptsize
{
"
红 茶"为一种搭配,应该翻译为"blac
k tea"
}
\end{spacing}
}
;
}
%\visible<5->
\end{pgfonlayer}
...
...
@@ -228,27 +230,27 @@
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
%%%
引入短语翻译
\begin{frame}
{
引入更大的翻译单元
}
%%%
引入短语翻译
\begin{frame}
{
引入更大的翻译单元
}
\begin{itemize}
\item
简单的单词翻译似乎不行,咋办?
\visible
<2->
{
\alert
{
- 引入更大的翻译单元
}
}
\item
简单的单词翻译似乎不行,咋办?
\visible
<2->
{
\alert
{
- 引入更大的翻译单元
}}
\begin{minipage}
[t]
{
0.40
\linewidth
}
\begin{center}
{
\scriptsize
\begin{tabular}
{
l | l
}
\only
<1>
{
单词翻译表
}
\only
<2->
{
\alert
{
\sout
{
单词
}}
词串翻译表
}
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
红
$
\to
$
red
&
0.8
\\
红
$
\to
$
black
&
0.1
\\
茶
$
\to
$
tea
&
0.8
\\
\visible
<2->
{
我 喜欢
$
\to
$
I like
&
0.3
\\
我 喜欢
$
\to
$
I liked
&
0.2
\\
绿 茶
$
\to
$
green tea
&
0.5
\\
绿 茶
$
\to
$
the green tea
&
0.1
\\
红 茶
$
\to
$
black tea
&
0.6
\\
\only
<1>
{
单词翻译表
}
\only
<2->
{
\alert
{
\sout
{
单词
}}
词串翻译表
}
&
P
\\
\hline
我
$
\to
$
I
&
0.6
\\
喜欢
$
\to
$
like
&
0.3
\\
红
$
\to
$
red
&
0.8
\\
红
$
\to
$
black
&
0.1
\\
茶
$
\to
$
tea
&
0.8
\\
\visible
<2->
{
我 喜欢
$
\to
$
I like
&
0.3
\\
我 喜欢
$
\to
$
I liked
&
0.2
\\
绿 茶
$
\to
$
green tea
&
0.5
\\
绿 茶
$
\to
$
the green tea
&
0.1
\\
红 茶
$
\to
$
black tea
&
0.6
\\
...
&
}
\end{tabular}
}
...
...
@@ -262,10 +264,10 @@
\begin{scope}
\begin{scope}
{
\footnotesize
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=east] (s) at (s1.west)
{$
s
=
$}
;
}
\end{scope}
...
...
@@ -294,10 +296,10 @@
\begin{scope}
[yshift=-6em]
\begin{scope}
{
\footnotesize
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=west] (s1) at (0,0)
{
我
}
;
\node
[anchor=west] (s2) at ([xshift=1.0em]s1.east)
{
喜欢
}
;
\node
[anchor=west] (s3) at ([xshift=1.0em]s2.east)
{
\alert
{
红
}
}
;
\node
[anchor=west] (s4) at ([xshift=1.0em]s3.east)
{
茶
}
;
\node
[anchor=east] (s) at (s1.west)
{$
s
=
$}
;
}
\end{scope}
...
...
@@ -332,12 +334,12 @@
\end{center}
\end{minipage}
\item
<4->
\alert
{
优点
}
很多
\item
<4->
\alert
{
优点
}
很多
\begin{itemize}
\item
翻译时候可以考虑更大范围的上下文信息
\\
比如:"红茶"中的"红"如果和"茶"搭配
...
\item
更好的局部调序,比如:短语中有"的"字
$
\to
$
... of ... 结构
\item
更大范围的目标语连续词串的使用,有利于
$
n
$
-gram语言模型选择译文
\item
翻译时候可以考虑更大范围的上下文信息
\\
比如:"红茶"中的"红"如果和"茶"搭配 ...
\item
更好的局部调序,比如:短语中有"的"字
$
\to
$
... of ... 结构
\item
更大范围的目标语连续词串的使用,有利于
$
n
$
-gram语言模型选择译文
\end{itemize}
\end{itemize}
...
...
@@ -345,14 +347,14 @@
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
\section
{
翻译中为什么要使用短语
}
\section
{
翻译中为什么要使用短语
}
%%%------------------------------------------------------------------------------------------------------------
%%%
回顾基于词的翻译模型
\begin{frame}
{
回顾 - 基于词的机器翻译
}
%%%
回顾基于词的翻译模型
\begin{frame}
{
回顾 - 基于词的机器翻译
}
\begin{itemize}
\item
对每个
\alert
{
单词
}
的翻译进行(任意)组合
\item
对每个
\alert
{
单词
}
的翻译进行(任意)组合
\end{itemize}
\vspace
{
-2em
}
...
...
@@ -362,13 +364,13 @@
\begin{scope}
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}
}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}
}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}
}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}}
;
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}
}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}
}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}
}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}
}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}
}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词
):
}}}}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词):
}}
}}
;
\draw
[->,very thick,ublue] (s1.south) -- ([yshift=-0.7em]s1.south);
\draw
[->,very thick,ublue] (s2.south) -- ([yshift=-0.7em]s2.south);
...
...
@@ -484,7 +486,7 @@
\begin{scope}
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}}
;
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}
}
;
\visible
<2->
{
\draw
[->,ultra thick,red,line width=2pt,opacity=0.7] ([xshift=-0.5em]t13.west) -- ([xshift=0.8em]t13.east) -- ([xshift=-0.2em]t22.west) -- ([xshift=0.8em]t22.east) -- ([xshift=-0.2em]t31.west) -- ([xshift=0.8em]t31.east) -- ([xshift=-0.2em]t41.west) -- ([xshift=0.8em]t41.east) -- ([xshift=-0.2em]t52.west) -- ([xshift=1.2em]t52.east);
...
...
@@ -497,16 +499,16 @@
}
\invisible
{
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}}
;
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}
}
;
}
\visible
<2->
{
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词)
:
}}
;
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词):
}}
;
\draw
[->,ultra thick,red,line width=1.5pt,opacity=0.7] (wtranslabel.east) -- ([xshift=1em]wtranslabel.east);
}
\invisible
{
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语)
:
}}
;
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语):
}}
;
\draw
[->,ultra thick,ublue,line width=1.5pt,opacity=0.7] ([xshift=0.65em]ptranslabel.east) -- ([xshift=1.65em]ptranslabel.east);
}
...
...
@@ -518,11 +520,11 @@
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
%%%
基于短语的翻译模型
\begin{frame}
{
基于"短语"的机器翻译
}
%%%
基于短语的翻译模型
\begin{frame}
{
基于"短语"的机器翻译
}
\begin{itemize}
\item
对每个
\alert
{
单词
}
及
\alert
{
连续词串
}
的翻译进行(任意)组合
\item
对每个
\alert
{
单词
}
及
\alert
{
连续词串
}
的翻译进行(任意)组合
\end{itemize}
\vspace
{
-2em
}
...
...
@@ -532,13 +534,13 @@
\begin{scope}
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}
}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}
}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}
}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}}
;
\node
[anchor=west] (s1) at (0,0)
{
\textbf
{
我
}
}
;
\node
[anchor=west] (s2) at ([xshift=2em]s1.east)
{
\textbf
{
对
}
}
;
\node
[anchor=west] (s3) at ([xshift=2em]s2.east)
{
\textbf
{
你
}
}
;
\node
[anchor=west] (s4) at ([xshift=2em]s3.east)
{
\textbf
{
表示
}
}
;
\node
[anchor=west] (s5) at ([xshift=2em]s4.east)
{
\textbf
{
满意
}
}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词
):
}}}}
;
\node
[anchor=south west] (sentlabel) at ([yshift=-0.5em]s1.north west)
{
\scriptsize
{
\textbf
{
\alert
{
待翻译句子(已经分词):
}}
}}
;
\draw
[->,very thick,ublue] (s1.south) -- ([yshift=-0.7em]s1.south);
\draw
[->,very thick,ublue] (s2.south) -- ([yshift=-0.7em]s2.south);
...
...
@@ -654,7 +656,7 @@
\begin{scope}
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}}
;
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=0em,xshift=-0.5em]t11.north west) -- ([xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label2)
{
\footnotesize
{
\textbf
{
单词翻译
}}
}
;
\draw
[->,ultra thick,red,line width=2pt,opacity=0.7] ([xshift=-0.5em]t13.west) -- ([xshift=0.8em]t13.east) -- ([xshift=-0.2em]t22.west) -- ([xshift=0.8em]t22.east) -- ([xshift=-0.2em]t31.west) -- ([xshift=0.8em]t31.east) -- ([xshift=-0.2em]t41.west) -- ([xshift=0.8em]t41.east) -- ([xshift=-0.2em]t52.west) -- ([xshift=1.2em]t52.east);
...
...
@@ -667,14 +669,14 @@
}
\visible
<2->
{
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}}
;
\draw
[decorate,thick,decoration={brace,amplitude=5pt,mirror}]
([yshift=-0.2em,xshift=-0.5em]t13.south west) -- ([yshift=-6.3em,xshift=-0.5em]t13.south west) node [pos=0.5,left,xshift=1.0em,yshift=0.0em,text width=5em,align=left] (label3)
{
\footnotesize
{
\textbf
{
短语翻译
}}
}
;
}
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词)
:
}}
;
\node
[anchor=north west] (wtranslabel) at ([yshift=-4em]t15.south west)
{
\scriptsize
{
翻译路径(仅含有单词):
}}
;
\draw
[->,ultra thick,red,line width=1.5pt,opacity=0.7] (wtranslabel.east) -- ([xshift=1em]wtranslabel.east);
\visible
<3->
{
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语)
:
}}
;
\node
[anchor=north west] (ptranslabel) at ([yshift=-5.5em]t15.south west)
{
\scriptsize
{
翻译路径(含有短语):
}}
;
\draw
[->,ultra thick,ublue,line width=1.5pt,opacity=0.7] ([xshift=0.65em]ptranslabel.east) -- ([xshift=1.65em]ptranslabel.east);
}
...
...
Section06-Neural-Machine-Translation/section06-test.tex
查看文件 @
1e144093
...
...
@@ -145,170 +145,27 @@
\subsection
{
注意力机制
}
%%%------------------------------------------------------------------------------------------------------------
%%% 解码 -
beam search
\begin{frame}
{
推断 -
Beam Search
}
%%% 解码 -
长度惩罚和覆盖度
\begin{frame}
{
推断 -
其它特征
}
\begin{itemize}
\item
\textbf
{
Greedy Search
}
: 目标语每一个位置,输出层的Softmax可以得到所有单词的概率,然后选择一个概率最大单词输出,下一个位置的预测就基于这一步输出的单词
\item
\textbf
{
Beach Search
}
: 为了避免贪婪方法造成的错误累加,可以每次对
$
b
$
个单词进行扩展,而不是只使用一个单词,其中
$
b
$
称做束的宽度 - 这样可以搜索更多可能的译文
\item
直接用
$
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)
$
进行解码,面临两方面问题
\begin{itemize}
\item
对
$
\textrm
{
P
}
(
y
_
j|
\textbf
{
y
}_{
<j
}
,
\textbf
{
x
}
)
$
进行乘积会导致长句的概率很低
\item
模型本身并没有考虑每个源语言单词被使用的程度,比如一个单词可能会被翻译了很多``次''
\end{itemize}
\item
<2-> 因此,解码时会使用其它特征与
$
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)
$
一起组成模型得分
$
score
(
\textbf
{
y
}
,
\textbf
{
x
}
)
$
,
$
score
(
\textbf
{
y
}
,
\textbf
{
x
}
)
$
也作为beam search的排序依据
\begin{eqnarray}
score(
\textbf
{
y
}
,
\textbf
{
x
}
)
&
=
&
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)/
\textrm
{
lp
}
(
\textbf
{
y
}
) +
\textrm
{
cp
}
(
\textbf
{
y
}
,
\textbf
{
x
}
)
\nonumber
\\
\textrm
{
lp
}
(
\textbf
{
y
}
)
&
=
&
\frac
{
(5 + |
\textbf
{
y
}
|)
^
\alpha
}{
(5 + 1)
^
\alpha
}
\nonumber
\\
\textrm
{
cp
}
(
\textbf
{
y
}
,
\textbf
{
x
}
)
&
=
&
\beta
\cdot
\sum\nolimits
_{
i=1
}^{
|
\textbf
{
x
}
|
}
\log
(
\min
(
\sum\nolimits
_{
j
}^{
|
\textbf
{
y
}
|
}
a
_{
ij
}
, 1)))
\nonumber
\end{eqnarray}
\vspace
{
-0.5em
}
\begin{itemize}
\item
lp会惩罚译文过短的结果(长度惩罚);cp会惩罚把某些源语单词对应到很多目标语单词的情况(覆盖度),被覆盖的程度用
$
\sum\nolimits
_{
j
}^{
|
\textbf
{
y
}
|
}
a
_{
ij
}$
度量;
$
\alpha
$
和
$
\beta
$
是超参,需要经验性设置
\end{itemize}
\end{itemize}
\vspace
{
-0.3em
}
\visible
<2->
{
\begin{center}
\begin{tikzpicture}
\begin{scope}
\tikzstyle
{
rnnnode
}
= [minimum height=1.1em,minimum width=3.5em,inner sep=2pt,rounded corners=1pt,draw,fill=red!20];
\tikzstyle
{
wnode
}
= [minimum height=1.0em,minimum width=3.5em,inner sep=2pt,rounded corners=1pt,draw,fill=white];
\visible
<3->
{
\node
[rnnnode,anchor=west,fill=green!20] (t1) at (0,0)
{
\tiny
{$
e
_
y
()
$}}
;
}
\visible
<7->
{
\node
[rnnnode,anchor=west,fill=green!20] (t2) at ([xshift=2.2em]t1.east)
{
\tiny
{$
e
_
y
()
$
(
$
\times
3
$
)
}}
;
}
\visible
<8->
{
\node
[rnnnode,anchor=west,fill=green!20] (t3) at ([xshift=2.2em]t2.east)
{
\tiny
{$
e
_
y
()
$
(
$
\times
3
$
)
}}
;
\node
[anchor=west,inner sep=2pt] (t4) at ([xshift=0.3em]t3.east)
{
\tiny
{
...
}}
;
}
\visible
<3->
{
\node
[rnnnode,anchor=south] (s1) at ([yshift=1em]t1.north)
{
\tiny
{$
\textbf
{
s
}_
1
$}}
;
}
\visible
<7->
{
\node
[rnnnode,anchor=south] (s2) at ([yshift=1em]t2.north)
{
\tiny
{$
\textbf
{
s
}_
2
$
(
$
\times
3
$
)
}}
;
}
\visible
<8->
{
\node
[rnnnode,anchor=south] (s3) at ([yshift=1em]t3.north)
{
\tiny
{$
\textbf
{
s
}_
3
$
(
$
\times
3
$
)
}}
;
\node
[anchor=west,inner sep=2pt] (s4) at ([xshift=0.3em]s3.east)
{
\tiny
{
...
}}
;
}
\visible
<3->
{
\node
[rnnnode,anchor=south,fill=blue!20] (o1) at ([yshift=1em]s1.north)
{
\tiny
{
softmax
}}
;
}
\visible
<7->
{
\node
[rnnnode,anchor=south,fill=blue!20] (o2) at ([yshift=1em]s2.north)
{
\tiny
{
softmax (
$
\times
3
$
)
}}
;
}
\visible
<8->
{
\node
[rnnnode,anchor=south,fill=blue!20] (o3) at ([yshift=1em]s3.north)
{
\tiny
{
softmax (
$
\times
3
$
)
}}
;
\node
[anchor=west,inner sep=2pt] (o4) at ([xshift=0.3em]o3.east)
{
\tiny
{
...
}}
;
}
\node
[wnode,anchor=north] (wt1) at ([yshift=-0.8em]t1.south)
{
\tiny
{
EOS
}}
;
\visible
<6->
{
\node
[wnode,anchor=north] (wt2) at ([yshift=-0.8em]t2.south)
{
\tiny
{
Have
}}
;
\node
[wnode,anchor=north] (wt2copy1) at ([xshift=-0.2em,yshift=-0.2em]wt2.north)
{
\tiny
{
Have
}}
;
\node
[wnode,anchor=north] (wt2copy2) at ([xshift=-0.4em,yshift=-0.4em]wt2.north)
{
\tiny
{
Have
}}
;
}
\visible
<8->
{
\node
[wnode,anchor=north,inner sep=2pt] (wt3) at ([yshift=-0.8em]t3.south)
{
\tiny
{
you
}}
;
\node
[wnode,anchor=north] (wt3copy1) at ([xshift=-0.2em,yshift=-0.2em]wt3.north)
{
\tiny
{
you
}}
;
\node
[wnode,anchor=north] (wt3copy2) at ([xshift=-0.4em,yshift=-0.4em]wt3.north)
{
\tiny
{
you
}}
;
}
\visible
<5->
{
\node
[wnode,anchor=center,inner sep=2pt] (wo1) at ([xshift=0.4em,yshift=1.8em]o1.north)
{
\tiny
{
Have
}}
;
\node
[wnode,anchor=north] (wo1copy1) at ([xshift=-0.2em,yshift=-0.2em]wo1.north)
{
\tiny
{
Have
}}
;
\node
[wnode,anchor=north] (wo1copy2) at ([xshift=-0.4em,yshift=-0.4em]wo1.north)
{
\tiny
{
Have
}}
;
}
\visible
<8->
{
\node
[wnode,anchor=center,inner sep=2pt] (wo2) at ([xshift=0.4em,yshift=1.8em]o2.north)
{
\tiny
{
you
}}
;
\node
[wnode,anchor=north] (wo2copy1) at ([xshift=-0.2em,yshift=-0.2em]wo2.north)
{
\tiny
{
you
}}
;
\node
[wnode,anchor=north] (wo2copy2) at ([xshift=-0.4em,yshift=-0.4em]wo2.north)
{
\tiny
{
you
}}
;
}
\visible
<8->
{
\node
[wnode,anchor=center,inner sep=2pt] (wo3) at ([xshift=0.4em,yshift=1.8em]o3.north)
{
\tiny
{
learned
}}
;
\node
[wnode,anchor=north] (wo3copy1) at ([xshift=-0.2em,yshift=-0.2em]wo3.north)
{
\tiny
{
learned
}}
;
\node
[wnode,anchor=north] (wo3copy2) at ([xshift=-0.4em,yshift=-0.4em]wo3.north)
{
\tiny
{
learned
}}
;
}
\visible
<3->
{
\foreach
\x
in
{
1
}{
\draw
[->] ([yshift=-0.7em]t
\x
.south) -- ([yshift=-0.1em]t
\x
.south);
\draw
[->] ([yshift=0.1em]t
\x
.north) -- ([yshift=-0.1em]s
\x
.south);
\draw
[->] ([yshift=0.1em]s
\x
.north) -- ([yshift=-0.1em]o
\x
.south);
}
}
\visible
<5->
{
\draw
[->] ([yshift=0.1em]o1.north) -- ([yshift=0.8em]o1.north) node [pos=0.5,right]
{
\tiny
{
top-3
}}
;
}
\visible
<7->
{
\foreach
\x
in
{
2
}{
\draw
[->] ([yshift=-0.7em]t
\x
.south) -- ([yshift=-0.1em]t
\x
.south);
\draw
[->] ([yshift=0.1em]t
\x
.north) -- ([yshift=-0.1em]s
\x
.south);
\draw
[->] ([yshift=0.1em]s
\x
.north) -- ([yshift=-0.1em]o
\x
.south);
\draw
[->] ([yshift=0.1em]o
\x
.north) -- ([yshift=0.8em]o
\x
.north) node [pos=0.5,right]
{
\tiny
{
top-3
}}
;
}
}
\visible
<8->
{
\foreach
\x
in
{
3
}{
\draw
[->] ([yshift=-0.7em]t
\x
.south) -- ([yshift=-0.1em]t
\x
.south);
\draw
[->] ([yshift=0.1em]t
\x
.north) -- ([yshift=-0.1em]s
\x
.south);
\draw
[->] ([yshift=0.1em]s
\x
.north) -- ([yshift=-0.1em]o
\x
.south);
\draw
[->] ([yshift=0.1em]o
\x
.north) -- ([yshift=0.8em]o
\x
.north) node [pos=0.5,right]
{
\tiny
{
top-3
}}
;
}
}
\visible
<3->
{
\draw
[->] ([xshift=-0.5em]s1.west) -- ([xshift=-0.1em]s1.west) node [pos=0,left,inner sep=1pt]
{
\tiny
{
0
}}
;
}
\visible
<7->
{
\draw
[->] ([xshift=0.1em]s1.east) -- ([xshift=-0.1em]s2.west);
}
\visible
<8->
{
\draw
[->] ([xshift=0.1em]s2.east) -- ([xshift=-0.1em]s3.west);
}
\visible
<6->
{
\draw
[->,very thick,dotted] (wo1.east) .. controls +(east:0.6) and +(west:0.8) ..(wt2copy2.west);
}
\visible
<8->
{
\draw
[->,very thick,dotted] (wo2.east) .. controls +(east:0.6) and +(west:0.8) ..(wt3copy2.west);
}
\visible
<7->
{
\node
[circle,draw,anchor=north,inner sep=2pt,fill=orange!20] (c2) at ([yshift=-2.5em]t1.south)
{
\tiny
{$
\textbf
{
C
}_
2
$}}
;
\node
[circle,draw,inner sep=2pt,fill=orange!20] (c2copy1) at ([yshift=-0.1em,xshift=-0.1em]c2)
{
\tiny
{$
\textbf
{
C
}_
2
$}}
;
\node
[circle,draw,inner sep=2pt,fill=orange!20] (c2copy2) at ([yshift=-0.2em,xshift=-0.2em]c2)
{
\tiny
{$
\textbf
{
C
}_
2
$}}
;
\draw
[->] ([xshift=-0.9em]c2.west) -- ([xshift=-0.3em]c2.west);
\draw
[->] ([xshift=0.1em]c2.east) .. controls +(east:1.5) and +(west:0.8) ..([yshift=-0.3em,xshift=-0.1em]s2.west);
}
\visible
<8->
{
\node
[circle,draw,anchor=north,inner sep=2pt,fill=orange!20] (c3) at ([yshift=-2.5em]t2.south)
{
\tiny
{$
\textbf
{
C
}_
3
$}}
;
\node
[circle,draw,inner sep=2pt,fill=orange!20] (c3copy1) at ([yshift=-0.1em,xshift=-0.1em]c3)
{
\tiny
{$
\textbf
{
C
}_
3
$}}
;
\node
[circle,draw,inner sep=2pt,fill=orange!20] (c3copy2) at ([yshift=-0.2em,xshift=-0.2em]c3)
{
\tiny
{$
\textbf
{
C
}_
3
$}}
;
\draw
[->] ([xshift=-0.9em]c3.west) -- ([xshift=-0.3em]c3.west);
\draw
[->] ([xshift=0.1em]c3.east) .. controls +(east:1.5) and +(west:0.8) ..([yshift=-0.3em,xshift=-0.1em]s3.west);
}
\visible
<3->
{
\node
[anchor=east] (vocab) at ([xshift=-5em]s1.west)
{
\tiny
{$
\begin
{
bmatrix
}
\textrm
{
Have
}
&
0
.
50
\\
\textrm
{
I
}
&
0
.
02
\\
\textrm
{
it
}
&
0
.
03
\\
\textrm
{
has
}
&
0
.
30
\\
\textrm
{
you
}
&
0
.
01
\\
\textrm
{
the
}
&
0
.
01
\\
\textrm
{
a
}
&
0
.
01
\\
\textrm
{
an
}
&
0
.
02
\\
\textrm
{
he
}
&
0
.
03
\\
\textrm
{
she
}
&
0
.
01
\\
\textrm
{
are
}
&
0
.
00
\\
\textrm
{
am
}
&
0
.
01
\\
...
&
...
\end
{
bmatrix
}$}}
;
\node
[anchor=south] (vocablabel) at (vocab.north)
{
\tiny
{
单词的概率分布
}}
;
\draw
[->,red,very thick,dotted] (o1.west) .. controls +(west:1) and +(east:2) .. ([yshift=1em]vocab.south east);
}
\visible
<4->
{
\node
[anchor=east,inner sep=1pt] (vocabtopn) at ([xshift=-0.5em,yshift=-0.5em]wo1.west)
{
\tiny
{$
\begin
{
bmatrix
}
\textrm
{
Have
}
\\
\textrm
{
has
}
\\
\textrm
{
it
}
\end
{
bmatrix
}$}}
;
\draw
[->] ([yshift=-1.6em,xshift=-0.4em]vocab.north east) .. controls +(east:1) and +(west:1) .. ([xshift=0.1em,yshift=0.4em]vocabtopn.west) node [pos=0.3,below] (topnlabel)
{
\tiny
{
top-3
}}
;
\visible
<4->
{
\node
[anchor=north] (cap) at (vocab.south east)
{
\scriptsize
{
\textbf
{
束搜索(
$
b
=
3
$
)
}}}
;
}
}
\end{scope}
\end{tikzpicture}
\end{center}
}
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
...
...
Section06-Neural-Machine-Translation/section06.tex
查看文件 @
1e144093
...
...
@@ -2438,6 +2438,30 @@ $\textrm{``you''} = \argmax_{y} \textrm{P}(y|\textbf{s}_1, \alert{\textbf{C}})$
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
%%% 解码 - 长度惩罚和覆盖度
\begin{frame}
{
推断 - 其它特征
}
\begin{itemize}
\item
直接用
$
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)
$
进行解码,面临两方面问题
\begin{itemize}
\item
对
$
\textrm
{
P
}
(
y
_
j|
\textbf
{
y
}_{
<j
}
,
\textbf
{
x
}
)
$
进行乘积会导致长句的概率很低
\item
模型本身并没有考虑每个源语言单词被使用的程度,比如一个单词可能会被翻译了很多``次''
\end{itemize}
\item
<2-> 因此,解码时会使用其它特征与
$
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)
$
一起组成模型得分
$
score
(
\textbf
{
y
}
,
\textbf
{
x
}
)
$
,
$
score
(
\textbf
{
y
}
,
\textbf
{
x
}
)
$
也作为beam search的排序依据
\begin{eqnarray}
score(
\textbf
{
y
}
,
\textbf
{
x
}
)
&
=
&
\textrm
{
P
}
(
\textbf
{
y
}
|
\textbf
{
x
}
)/
\textrm
{
lp
}
(
\textbf
{
y
}
) +
\textrm
{
cp
}
(
\textbf
{
y
}
,
\textbf
{
x
}
)
\nonumber
\\
\textrm
{
lp
}
(
\textbf
{
y
}
)
&
=
&
\frac
{
(5 + |
\textbf
{
y
}
|)
^
\alpha
}{
(5 + 1)
^
\alpha
}
\nonumber
\\
\textrm
{
cp
}
(
\textbf
{
y
}
,
\textbf
{
x
}
)
&
=
&
\beta
\cdot
\sum\nolimits
_{
i=1
}^{
|
\textbf
{
x
}
|
}
\log
(
\min
(
\sum\nolimits
_{
j
}^{
|
\textbf
{
y
}
|
}
a
_{
ij
}
, 1)))
\nonumber
\end{eqnarray}
\vspace
{
-0.5em
}
\begin{itemize}
\item
lp会惩罚译文过短的结果(长度惩罚);cp会惩罚把某些源语单词对应到很多目标语单词的情况(覆盖度),被覆盖的程度用
$
\sum\nolimits
_{
j
}^{
|
\textbf
{
y
}
|
}
a
_{
ij
}$
度量;
$
\alpha
$
和
$
\beta
$
是超参,需要经验性设置
\end{itemize}
\end{itemize}
\end{frame}
%%%------------------------------------------------------------------------------------------------------------
%%% 实验结果
\begin{frame}
{
效果
}
%% 实用注意力机制带来的提升
...
...
编写
预览
Markdown
格式
0%
重试
或
添加新文件
添加附件
取消
您添加了
0
人
到此讨论。请谨慎行事。
请先完成此评论的编辑!
取消
请
注册
或者
登录
后发表评论