- Download the ISO image from http://sslmit.unibo.it/~baroni/knorpora_pages/knorpora.html.
- Burn the image to a CD.
- Put the CD in the CD drive of a computer meeting the following requirements:
Boot from CD (in most cases, this means: start the computer -- if
it does not work, change the boot order in the bios).
You will probably get better performance if you type knoppix
dma at the boot prompt.
Once the KDE starts (it will take a while -- you are launching a
full fledged operating system from a CD...), if you plan to go online
(e.g., to download more data to play with), click on the penguin icon
at the bottom left, select the appropriate connection type from the
Network/Internet menu, and configure it (you will need to collect the
configuration data from the computer or from an administrator before you
boot from CD).
If you never used a Unix-like operating systems, you should
probably learn something about it -- see, for example, my Unix
links for some tutorials and the like.
There are various ways to save your data and configurations with
Knoppix, and, thus, Knorpora (although none of them is completely
trivial). A good place to start gathering information about this and
other issues is http://www.knoppix.net/
- Intel-compatible CPU, i486 minimum.
- 96MB of RAM.
- Bootable CD-ROM drive.
Getting started with the Knorpora-specific stuff
There is nothing special to do -- just read the documentation
pertaining to the various tools and data.
I added all the Knorpora-specific paths to the PATH variable in the
.bashrc file, so in general you should be able to launch the
applications simply by typing their name at the prompt.
My policy regarding where to put what has been the following:
- If the standard installation procedure of a tool already specifies a default directory (e.g., /usr/local/bin), I did not change it.
- Everything else goes to /usr/local/knorpora.
To test the tools that work on English data, you can copy some of
the archives that come with the Natural Language Toolkit to the
knoppix home and unzip them.
Getting acquainted with the Knorpora tools
- To get started with WordNet type man wn at the prompt.
- To get started with the Natural Language Toolkit and the data that come with it take a look at /usr/local/knorpora/nltk-1.3/README.first.
- To get started with ACOPOST, take a look at /usr/local/knorpora/acopost-1.8.4/doc/ug and to /usr/local/knorpora/acopost-1.8.4/ItaACOPOSTModels/Readme.ItaACOPOSTModels for information about the pre-trained Italian models.
- To get started with the TreeTagger, take a look at the README and README.script files in the /usr/local/knorpora/treetagger folder.
- You can get started with the fnTBL Toolkit by looking at the README file in /usr/local/knorpora/fnTBL-1.1.
- To get started with FreeLing, take a look at the documentation in /usr/local/share/FreeLing/doc.
- Documentation about ChaSen and Ipadic can be found in the directories /usr/share/doc/chasen and /usr/share/doc/ipadic, respectively.
- You can get started with the BootCaT tools by looking at the Readme.BootCaT-0.1.2 file in /usr/local/knorpora/BootCaT-0.1.2.
- For K-vec++, take a look at the Readme.txt file in /usr/local/knorpora/K-vec++.
- For nsp, type perldoc /usr/local/knorpora/nsp-v0.69/Docs/README.pod.
- For UCS, take a look at the README file in /usr/local/knorpora/UCS.
- For SenseClusters, type perldoc /usr/local/knorpora/SenseClusters-v0.49/README.Intro.pod.
- To get started with the Bow Toolkit take a look at the information on the Bow Toolkit webpage and/or type rainbow --help | less, crossbow --help | less, arrow --help | less at the prompt.
- For Jan Daciuk's Finite State Utilities, take a look at the man page for the utility/ies you are interested in (e.g., man fsa_build, man tr_morph, etc.)
- To see the usage of kwic, just type kwic at the prompt.
- To get started with R, type R at the prompt and then help.start().
- Use perldoc to read documentation about the perl modules (to get started with the handy similarity.pl utility that comes with the WordNet::Similarity module type similarity.pl --help | less at the prompt).
For most tools and data, you can also find documentation at the relevant site on the net (see the links in the Knorpora Software page).
Back to the Welcome to Knorpora page.