1.Overview of Course, Parsing and Statistical Parsing [ppt] [pdf]

Motivation for statistical parsing, recursive phrase structure. Attachment decisions and probabilities. The Penn Treebank. Top-down, and bottom-up parsing.

• Introductory background reading: Eugene Charniak. 1997. Statistical techniques for natural language parsing. AI Magazine.
• Jurafsky and Martin, 12.0-12.3, if you don't know much syntax.
• Jurafsky and Martin, 12.4, for treebanks.
• Jurafsky and Martin, 13.0-13.3 for parsing as search

2.PCFGs and the CKY algorithm [ppt] [pdf]

PCFGs. Grammar transformations. Recursive parsing and memoization. Dynamic programming for parsing: the CKY algorithm.

• Jurafsky and Martin, sections 13.4.0-13.4.1 CKY parsing
• Jurafsky and Martin, sections 14.0-14.2, PCFGs and probabilistic CKY parsing

3.Generalized CKY parsing and unlexicalized parsing [ppt] [pdf]

Unaries and empties. Parser evaluation. Improving the context-freedom assumptions of grammars: accurate unlexicalized parsing.

• Dan Klein and Christopher D. Manning. 2001. Parsing with Treebank Grammars: Empirical Bounds, Theoretical Models, and the Structure of the Penn Treebank. Proceedings of the 39th Annual Meeting of the ACL, pp. 330-337.
• Dan Klein and Christopher D. Manning. 2003. Accurate Unlexicalized Parsing. ACL 2003, pp. 423-430.

4.Search in parsing and lexicalized probabilistic parsing [ppt] [pdf]

Beam parsing. Agenda-based (chart) parsing. A* parsing. Lexicalized probabilistic context-free grammars: The Charniak (1997) model.

• Dan Klein and Christopher D. Manning. 2001. Parsing and Hypergraphs. Proceedings of the 7th International Workshop on Parsing Technologies (IWPT-2001), pp. 123-134.
• Dan Klein and Christopher D. Manning. 2003. A* Parsing: Fast Exact Viterbi Parse Selection. HLT-NAACL 2003.
• Sharon A. Caraballo and Eugene Charniak. 1998. New Figures of Merit for Best-First Probabilistic Chart Parsing. Computational Linguistics 24: 275-298.
• Slav Petrov and Dan Klein. 2007. Improved Inference for Unlexicalized Parsing. North American Chapter of the Association for Computational Linguistics (HLT-NAACL 2007).
• Eugene Charniak. 1997. Statistical parsing with a context-free grammar and word statistics. Proceedings of the Fourteenth National Conference on Artificial Intelligence (AAAI 1997).
   

5.Treebanks and statistical parsing [ppt] [pdf]

Lexicalized parsing: Collins (1997/1999). The status of information in treebanks.

• Michael Collins. 1997. Three Generative, Lexicalised Models for Statistical Parsing. Proceedings of the 35th Annual Meeting of the ACL, Madrid.
• Michael Collins. 2003. Head-Driven Statistical Models for Natural Language Parsing
• Daniel M. Bikel. 2004. A Distributional Analysis of a Lexicalized Statistical Parsing Model. EMNLP 2004.

6.Multilingual parsing and dependency parsing [ppt] [pdf]

• Roger Levy and Christopher D. Manning. 2003. Is it harder to parse Chinese, or the Chinese Treebank?. ACL 2003, pp. 439-446.
• Eisner, Jason and Giorgio Satta (1999). Efficient parsing for bilexical context-free grammars and head automaton grammars. Proceedings of the 37th Annual Meeting of the Association for Computational Linguistics, pages 457-464.

7.Discriminative parsing [ppt] [pdf]

An introduction to discriminative parsing. Features in discriminative parsers, presented using some of Mark Johnson's slides on discriminative reranking.

• Michael Collins. 2004. Parameter Estimation for Statistical Parsing Models: Theory and Practice of Distribution-Free Methods. In Harry Bunt, John Carroll and Giorgio Satta, editors, New Developments in Parsing Technology, Kluwer.
• Ryan McDonald, Koby Crammer and Fernando Pereira. 2005. Online Large-Margin Training of Dependency Parsers. 43rd Annual Meeting of the Association for Computational Linguistics.
• Eugene Charniak and Mark Johnson. 2005. Coarse-to-Fine n-Best Parsing and MaxEnt Discriminative Reranking 43rd Annual Meeting of the Association for Computational Linguistics.

위대한 ‘발견’ 이라는 것은 그 발견 자체를 목적으로 일어나는 경우는 극히 드물다.
나 와 Zinkernagel 박사가 찾아낸 MHC restriction 이라는 이론 또한 “T 세포가 체내에서 어떠한 방법으로 바이러스 (또는 병균)을 인식하고 면역 반응을 일으킬까?” 하는 간단한 질문에서 시작되었다.

내가 면역학과 바이러스학에 흥미를 갖기 시작한 것은 University of Queensland 에서 수의학을 전공하면서였고, 본격적으로 ‘바이러스 감염에 의한 면역반응’ 이라는 주제로 연구를 시작한 것은 the John Curtin School of Medical Research 에서 Post doc.으로 있을 때부터였다.
곳에서 노벨상 공동수상자인 Zinkernagel 박사와 killer T 세포의 활동에 대한 연구를 시작한 나는 다른 과학자들과는 약간 다른 관점에서 연구를 시작하였다. 가설을 세우고 그것을 증명하기 위해 연구하는 기존의 과학자들과는 달리, 나는 내 가설이 ‘거짓’임을 입증하는데 초점을 두고, 그 문제점을 찾으려고 노력하였다. 내 생각에 발견이란 하나의 가설을 세우고, 그 가설의 문제점을 찾아내고, 그 문제점이 암시해주는 ‘흥미로운 단서’를 붙잡아 또 다른 가설로 시작하는, 이러한 과정의 반복으로 이루어지는 노력의 과실이라고 생각한다. 내가 졸업한 John Curtin School 의 director 인 Kevin Lafferty 박사는 나의 노벨상 수상을 ‘호기심이 이끌어낸 연구의 승리’ 라고 칭할 만큼 나는 내 연구에 많은 흥미와 열정, 노력을 기울였다.

성공적인 과학자가 가져야 할 태도는 간단명료하다. 언제나 자신의 논리와 이론에 믿음을 갖고 고수하되, 그것이 틀릴 경우 과감히 놓아버리고 열린 눈과 마음으로 새로운 사실을 받아들이고 호기심이 이끄는 데로 나아간다면 누구나 과학자로서 큰 성공을 할 수 있는 날이 올 것이다. 

Linux / Unix Command: ps / kill

Command Library

Even on Linux it sometimes happens that processes wear out their welcome and stick around longer than you would like them to. They simply ignore your request to close up and go away. Fortunately you have a powerful gun at your disposal that will clean out anything that doesn't get a hint: The Arnold Schwarzenegger equivalent among the Linux commands is the kill command.

The kill command works together with the ps command

• With the ps command (ps stands for "process status") you find out the identity of the program you want to get rid of.
• Then kill will finish it off.

  Here's how you do it:

The ps Command

It is used to display the currently running processes on Unix/Linux systems. If you know the 'Task-Manager" which pops up under Windows NT/2000/XP when you press CTRL+ALT+DEL then you will quickly grasp what ps does on Linux. It can show you the processes on your system in various formats. Here is a typical example:

To display all processes owned by the current user type ps ux and hit return:

$ ps ux

This will show a listing of processes similar to:

USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
jhaas 3064 0.1 3.6 18324 9088 ? S 17:55 0:00 /usr/bin/gnome-session
jhaas 3107 0.0 0.3 3128 968 ? S 17:55 0:00 /usr/bin/ssh-agent /etc/X11/xinit/Xclients
jhaas 3112 0.1 2.9 11208 7376 ? S 17:55 0:00 /usr/libexec/gconfd-2 11
jhaas 3130 0.0 0.6 3584 1696 ? S 17:55 0:00 xscreensaver -nosplash
jhaas 3133 0.8 5.3 21744 13216 ? S 17:55 0:03 gnome-panel --sm-client-id default2
jhaas 3137 0.0 2.4 16296 6172 ? S 17:55 0:00 magicdev --sm-client-id default4
jhaas 3141 0.0 2.8 16676 6936 ? S 17:55 0:00 eggcups --sm-client-id default6
jhaas 3143 0.0 1.7 11652 4308 ? S 17:55 0:00 pam-panel-icon --sm-client-id default0
jhaas 3145 0.2 5.4 24280 13564 ? S 17:55 0:01 /usr/bin/python /usr/bin/rhn-applet-gui --sm-client-id default5
jhaas 3154 0.3 4.1 23856 10172 ? S 17:55 0:01 evolution
jhaas 3157 0.0 0.9 5636 2228 ? S 17:55 0:00 oafd --ac-activate --ior-output-fd=10
jhaas 3216 0.2 3.1 12788 7844 ? S 17:56 0:00 emacs
jhaas 3218 1.3 4.0 20248 9904 ? R 17:56 0:04 gnome-terminal
jhaas 3219 0.0 0.2 1852 572 ? S 17:56 0:00 [gnome-pty-helpe]
jhaas 3220 0.0 0.5 4316 1388 pts/0 S 17:56 0:00 bash
jhaas 3268 0.0 0.2 2636 684 pts/0 R 18:01 0:00 ps ux

The kill Command

Now, if you want to terminate for example the emacs process you would look up the process identifier (PID) in the above table (3216), and say:

$ kill -9 3216

The -9 will ensure "execution".

A convenient short cut is the Alt-Ctrl-Esc key combination, which allows you to simply click on the application you want to kill.

df - Linux command line tool to display disk space usage

-h, --human-readable
print sizes in human readable format (e.g., 1K 234M 2G)

-H, --si
likewise, but use powers of 1000 not 1024

-i, --inodes
list inode information instead of block usage

-k like --block-size=1K

-l, --local
limit listing to local file systems

--no-sync
do not invoke sync before getting usage info (default)

-P, --portability
use the POSIX output format

--sync invoke sync before getting usage info

-t, --type=TYPE
limit listing to file systems of type TYPE

-T, --print-type
print file system type

-x, --exclude-type=TYPE
limit listing to file systems not of type TYPE

-v (ignored)

--help display this help and exit

--version
output version information and exit

Examples

If you use df with no options it will display all the available filesystems' info

$ df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 151225248 16280980 127262392 12% /
tmpfs 517924 0 517924 0% /lib/init/rw
udev 10240 60 10180 1% /dev
tmpfs 517924 0 517924 0% /dev/shm
/dev/sdb1 76896316 54820876 18169240 76% /media/discob
sshfs#ggarron@alketech.com:/home/ggarron/
7999999992 0 7999999992 0% /mnt/remote

To better understand what is in the screen can be used the -h (human readable)

$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 145G 16G 122G 12% /
tmpfs 506M 0 506M 0% /lib/init/rw
udev 10M 60K 10M 1% /dev
tmpfs 506M 0 506M 0% /dev/shm
/dev/sdb1 74G 53G 18G 76% /media/discob
sshfs#ggarron@alketech.com:/home/ggarron/
7.5T 0 7.5T 0% /mnt/remote

To display only local mounted systemfilesystems use the -l (local) options, you will see the sshfs remote filesystem is not shown anymore.

$ df -hl
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 145G 16G 122G 12% /
tmpfs 506M 0 506M 0% /lib/init/rw
udev 10M 60K 10M 1% /dev
tmpfs 506M 0 506M 0% /dev/shm
/dev/sdb1 74G 53G 18G 76% /media/discob

You can use it pointed to the device path or to the path where the filesystem is mounted

$ df -h /dev/sdb1
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 74G 53G 18G 76% /media/discob

$ df -h /media/discob/
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 74G 53G 18G 76% /media/discob

An A-Z Index of the Linux BASH command line

alias    Create an alias
apropos  Search Help manual pages (man -k)
awk      Find and Replace text, database sort/validate/index
break    Exit from a loop
builtin  Run a shell builtin
bzip2    Compress or decompress named file(s)

cal      Display a calendar
case     Conditionally perform a command
cat      Display the contents of a file
cd       Change Directory
cfdisk   Partition table manipulator for Linux
chgrp    Change group ownership
chmod    Change access permissions
chown    Change file owner and group
chroot   Run a command with a different root directory
cksum    Print CRC checksum and byte counts
clear    Clear terminal screen
cmp      Compare two files
comm     Compare two sorted files line by line
command  Run a command - ignoring shell functions
continue Resume the next iteration of a loop
cp       Copy one or more files to another location
cron     Daemon to execute scheduled commands
crontab  Schedule a command to run at a later time
csplit   Split a file into context-determined pieces
cut      Divide a file into several parts

date     Display or change the date & time
dc       Desk Calculator
dd       Data Dump - Convert and copy a file
declare  Declare variables and give them attributes
df       Display free disk space
diff     Display the differences between two files
diff3    Show differences among three files
dig      DNS lookup
dir      Briefly list directory contents
dircolors Colour setup for `ls'
dirname  Convert a full pathname to just a path
dirs     Display list of remembered directories
du       Estimate file space usage

echo     Display message on screen
egrep    Search file(s) for lines that match an extended expression
eject    Eject removable media
enable   Enable and disable builtin shell commands
env      Environment variables
ethtool  Ethernet card settings
eval     Evaluate several commands/arguments
exec     Execute a command
exit     Exit the shell
expand   Convert tabs to spaces
export   Set an environment variable
expr     Evaluate expressions

false    Do nothing, unsuccessfully
fdformat Low-level format a floppy disk
fdisk    Partition table manipulator for Linux
fgrep    Search file(s) for lines that match a fixed string
file     Determine file type
find     Search for files that meet a desired criteria
fmt      Reformat paragraph text
fold     Wrap text to fit a specified width.
for      Expand words, and execute commands
format   Format disks or tapes
free     Display memory usage
fsck     File system consistency check and repair
ftp      File Transfer Protocol
function Define Function Macros

gawk     Find and Replace text within file(s)
getopts  Parse positional parameters
grep     Search file(s) for lines that match a given pattern
groups   Print group names a user is in
gzip     Compress or decompress named file(s)

hash     Remember the full pathname of a name argument
head     Output the first part of file(s)
history  Command History
hostname Print or set system name

id       Print user and group id's
if       Conditionally perform a command
ifconfig Configure a network interface
import   Capture an X server screen and save the image to file
install  Copy files and set attributes

join     Join lines on a common field

kill     Stop a process from running

less     Display output one screen at a time
let      Perform arithmetic on shell variables
ln       Make links between files
local    Create variables
locate   Find files
logname  Print current login name
logout   Exit a login shell
look     Display lines beginning with a given string
lpc      Line printer control program
lpr      Off line print
lprint   Print a file
lprintd  Abort a print job
lprintq  List the print queue
lprm     Remove jobs from the print queue
ls       List information about file(s)
lsof     List open files

make     Recompile a group of programs
man      Help manual
mkdir    Create new folder(s)
mkfifo   Make FIFOs (named pipes)
mkisofs  Create an hybrid ISO9660/JOLIET/HFS filesystem
mknod    Make block or character special files
more     Display output one screen at a time
mount    Mount a file system
mtools   Manipulate MS-DOS files
mv       Move or rename files or directories

netstat  Networking information
nice     Set the priority of a command or job
nl       Number lines and write files
nohup    Run a command immune to hangups
nslookup Query Internet name servers interactively

passwd   Modify a user password
paste    Merge lines of files
pathchk  Check file name portability
ping     Test a network connection
popd     Restore the previous value of the current directory
pr       Prepare files for printing
printcap Printer capability database
printenv Print environment variables
printf   Format and print data
ps       Process status
pushd    Save and then change the current directory
pwd      Print Working Directory

quota    Display disk usage and limits
quotacheck Scan a file system for disk usage
quotactl Set disk quotas

ram      ram disk device
rcp      Copy files between two machines.
read     read a line from standard input
readonly Mark variables/functions as readonly
remsync  Synchronize remote files via email
return   Exit a shell function
rm       Remove files
rmdir    Remove folder(s)
rsync    Remote file copy (Synchronize file trees)

screen   Terminal window manager
scp      Secure copy (remote file copy)
sdiff    Merge two files interactively
sed      Stream Editor
select   Accept keyboard input
seq      Print numeric sequences
set      Manipulate shell variables and functions
sftp     Secure File Transfer Program
shift    Shift positional parameters
shopt    Shell Options
shutdown Shutdown or restart linux
sleep    Delay for a specified time
sort     Sort text files
source   Run commands from a file `.'
split    Split a file into fixed-size pieces
ssh      Secure Shell client (remote login program)
strace   Trace system calls and signals
su       Substitute user identity
sum      Print a checksum for a file
symlink  Make a new name for a file
sync     Synchronize data on disk with memory

tail     Output the last part of files
tar      Tape ARchiver
tee      Redirect output to multiple files
test     Evaluate a conditional expression
time     Measure Program running time
times    User and system times
touch    Change file timestamps
top      List processes running on the system
traceroute Trace Route to Host
trap     Run a command when a signal is set(bourne)
tr       Translate, squeeze, and/or delete characters
true     Do nothing, successfully
tsort    Topological sort
tty      Print filename of terminal on stdin
type     Describe a command

ulimit   Limit user resources
umask    Users file creation mask
umount   Unmount a device
unalias  Remove an alias
uname    Print system information
unexpand Convert spaces to tabs
uniq     Uniquify files
units    Convert units from one scale to another
unset    Remove variable or function names
unshar   Unpack shell archive scripts
until    Execute commands (until error)
useradd  Create new user account
usermod  Modify user account
users    List users currently logged in
uuencode Encode a binary file 
uudecode Decode a file created by uuencode

v        Verbosely list directory contents (`ls -l -b')
vdir     Verbosely list directory contents (`ls -l -b')
vi       Text Editor

watch    Execute/display a program periodically
wc       Print byte, word, and line counts
whereis  Report all known instances of a command    
which    Locate a program file in the user's path. 
while    Execute commands
who      Print all usernames currently logged in
whoami   Print the current user id and name (`id -un')
Wget     Retrieve web pages or files via HTTP, HTTPS or FTP

xargs    Execute utility, passing constructed argument list(s)
yes      Print a string until interrupted

.period  Run commands from a file
###      Comment / Remark