Meaning representation in natural language processing
Date of Issue2014
School of Computer Engineering
In this report, the semantic information in parse selection is analyzed. A Python software model was used to carry out feature engineering on semantic parsing results by parsers. The data used was from the SemCor corpus and WeScience corpus. Different types of semantic features were generated using the model and training and testing was conducted using a maximum entropy model TADM. Error analysis was performed on the entire SemCor and WeScience corpus by reproducing the old results. Generalized features provide better parse selection accuracy than more specific features. Further Machine learning was performed using ELM, Extreme Machine Learning technique, to compare the parse ranking results with TADM. The key fundamental task is to understand the meaning of a word in a sentence and semantic relations between words, resolving ambiguities by considering context.
DRNTU::Engineering::Computer science and engineering
Final Year Project (FYP)
Nanyang Technological University