The course Natural Language Generation has been cancelled.
It will be offered again in the next semester (Winter Semester 2015/16).
Apart from being a really good body guard, your new personal assistant robot Amanda can clean carpets, and exterminate mice. But to help her do a good job, you need to communicate about the situation. Amanda will have to describe rooms, answer questions such as “where is the mouse?” and report what she is doing.
In this seminar, we will look at the processing steps Amanda needs to make in order to communicate these things to you. You will get an overview of all steps of natural language generation, with a focus on situated interaction: conceptualizing sensory data, selecting information, planning an utterance, generating text and synthesizing speech.
For each step, there will be an introduction accompanied by pen and paper or programming exercises in class, and as homework assignments. At the end of the course, you will have a complete natural language generation system that can generate utterances for a restricted example scenario using existing NLG resources.
You will also present an NLG-related research paper and design your own extension for the NLG system based on that paper.