The Problem of Natural Language Processing

The only thing we have concentrated about right now is patent and scientific papers. The trick is which information are you looking for inside the text? And this is what differentiates what we are doing in my group than in probably in other groups. For our blog, we have not been that far to annotate plastic manufacturing text. This would be an ideal situation if you give me 10 artwork engineers to annotate 2000 texts of plastic manufacturing. We will find experts. You know, understanding a plastic text could also be done with a master student or somebody that is not an experienced engineer. Of course, it might not be the same quality as an expert, but maybe sufficient to understand what we need to