A few days ago I started a new NLP project. The goal of the project is to be able to automatically evaluate and classify content of texts, which are collected in the context of larger surveys. There are two different levels of complexity that I see in this problem. Complexity