Lexical Annotation Workbench - Readme This document contains the following:

• 1. A brief description of the Lexical Annotation Workbench
• 2. Installation Instructions
• 3. Samples
• 4. Configuring support for a tagset
• 5. Known Issues

1. What is it?

Lexical Annotation Workbench is an integrated environment for morphological annotation. It supports simple morphological annotation (assigning a lemma and tag to a word), integration and comparison of different annotations of the same text, searching for particular word, tag etc. The workbench runs on all operating systems supporting Java, including Windows and Linux. It is an open system extensible via plugins - e.g., views, import/export filters, helps. The primary file format is the PDT 2.0 format (a PML instantiation); it is possible to import and export data in several other formats (see samples below).

The application is organized around layers of annotation. Typically, each morphological layer (an m-layer) contains information provided by a specific tool or process, e.g., a morphological analyzer, a tagger, an annotator. In the PDT 2.0 format, each layer corresponds to a single file.

See this page for more details.

Development of this application was or has been partially supported by:

• Academy of Sciences of the Czech Republic (project ID: 1ET101120413),
• Grant Agency Czech Republic (project ID: P406/10/P328) and
• U.S. NSF (award #: 0916280).

2. Installation Instructions

Make sure you have Java Runtime Environment Version 6 (aka 1.6) installed. You can use this online test to determine it. If not, it can be downloaded here.

2.1. Windows

1. Unpack the installation archive to a directory of your choice (below, I assume it is the usual place C:\Program Files).
2. Make a shortcut on your desktop to javaw -Xmx196m -jar "C:/Program Files/Law/Law.jar" (do not forget the quotes). (Right click on the desktop, select New > Shortcut)
   Note: The shortcut can specify your login-name, say you are called john, then you can avoid the initial login dialog by using: javaw -Xmx196m -jar "C:/Program Files/Law/Law.jar" -u john. You can have multiple shortcuts, each for a different annotator.

See the advanced section below if you have less than 256MB memory or you want/need to change the default location of the configuration or logging files (e.g. because the application folder is not writable).

2.2. Linux

1. Unzip the installation archive into a directory of your choice, say /home/ab/law
2. Run the system as java -Xmx198m -jar /home/ab/law/Law.jar
   Note: The shortcut can specify your login-name, say you are called john, then you can avoid the initial login dialog by using: java -Xmx196m -jar /home/ab/law/Law.jar -u john.

See the advanced section below if you have less than 256MB memory or you want/need to change the default location of the configuration or logging files (e.g. because the application folder is not writable).

2.3. Advanced Installation Options

Changing memory configuration:

The number behind -Xmx says how much memory is available to the program, it should not be higher than 75% of your physical memory. For example, if you have only 128 MB RAM use -Xmx96m.

If you are working with large files and have enough memory you can increase the number.

Changing location of the configuration files

By default, the configuration files are in the conf subfolder of the application folder. You can change this (e.g. because you want to have multiple configurations) by using -confD folder switch:

• Windows: javaw -Xmx196m -jar "C:/Program Files/Law/Law.jar" -u john -confD D:/LawConf

Changing the location of the logging file

By default, problems with the program are logged to the problems.log file in the application folder. You can change this by using -logF file switch:

• Windows: javaw -Xmx196m -jar "C:/Program Files/Law/Law.jar" -logF C:/tmp/lawProblems.log

3. Samples

Sample data files are in the samples directory. Except the files in PDT 2.0 (PML) format, all other files must be imported selecting the proper format and encoding.

• sample0.m - sample in the PDT PML format (format used by the PDT 2.0 corpus), the native format of LAW. When this file is loaded, the sample0.w file is automatically loaded as well. The other possibility is to load the sample0.w file first and then load the sample0.m file.
• sample0.latin2.csts - csts format (format used by the PDT 1.0 corpus)
• sample0.latin2.tnt - tnt format
• sample0.utf8.tnt - the same file in utf8 encoding
• kafka2.wtl - format of Tree Tagger by Helmut Schmid, latin1 encoding.

4. Configuring support for a tagset

Law can display help for tags and possibly lemmas and provide a list of legal tags when a new tag is inserted. Which tagset should be used can be configured by editing the conf/law.Law.properties file. Help for several tagsets are provided in the tagset directory. In the future, we hope there will a more convenient way of configuring this.

5. Known Issues

• Initially, the LAW window layout is not very nice. Just drag the window borders to more reasonable locations.
• There is a memory leak in one of the underlying library - when a file is closed, it is not removed from the memory. After opening and closing several large files, LAW will crash. To avoid this simply restart the workbench before this can happen. If you have a large amount of memory you might consider to increase the memory available to LAW by changing the -Xmx startup parameter (see Changing memory Configuration above).