A Simple Model for Improving the Performance of the Stanford Parser for Action Detection in Textual Instructions

19602

Knowledge Triples