9The WSD4Kids demo implements a very simple Word Sense Disambiguation system and a graphical interface to interact with the system. The WSD system behind is based on a machine learning engine (Support Vector Machines) and uses a bag-of-word feature model.
The main idea is to show what is the main purpose of the WSD in a friendly way. We selected several Dutch words with clearly different meanings, and included one picture for every meaning of these words. You first need to teach the system, by providing examples (sentences or pieces of text) for the different meanings of the words, and selecting in each case the picture belonging to the specific meaning. The system learns how to identify different meanings through the examples provided together with the correct picture.
Once the system is trained, you can check how it works. You need to enter a sentence containing one of the the target words, with a specific meaning, and the system will automatically select the picture that belongs to the correct meaning (or that one that the system thinks is the most likely).
You can play with the demo (currently is only for Dutch text) here: http://kyoto.let.vu.nl/~izquierdo/wsd4kids/