WebLicht - A Web-Based Analysis Tool

WebLicht Logo

WebLicht is a web-based tool to semi-automatically annotate texts for linguistics and humanities research. It removes many of the time-consuming technical aspects of natural language processing - software installation, data format conversion, and error handling - so that users can immediately process their texts from any computer with a current web browser. Development on WebLicht started in October 2008 as part of CLARIN-D's predecessor project D-SPIN, and CLARIN-D is enhancing WebLicht to provide a uniform, distributed, accessible infrastructure for the largest possible research community.

WebLicht is a fully-functional virtual research environment for text studies. WebLicht provides a user-friendly web interface to construct and run custom processing chains of web services - each encapsulating some specific linguistic tool - to produce complete processing workflows. Processing a text to extract the most information from it is often a complex task where multiple tools, written in different programming languages for different environments with generally incompatible formats, must be run sequentially or can run in parallel but require users to collate the output. WebLicht handles this complex, low-level processing by wrapping all the tools as services it provides in a framework that uses a multilayered and multifaceted, XML-based format to encode all inputs and outputs.

WebLicht assigns each token an identifier that follows it through all processing steps, so that all added annotation, whether lexical, syntactic, or semantic, is directly connected to the corresponding element of text. Users have access to enhanced tools for visualizing annotated language data, including easy to use tables, tree-drawing suites, and enhancements for mapping locations, viewing search results in texts and annotations, and displaying segmented audio data.

You can login to WebLicht here if you have an account in the CLARIN Service Provider Federation. Detailed information can be found here.