This repo contains homeworks and assignments completed as part of CSCI-544 Aplied Natural Language Processing at USC in Spring 2023.
Homework 1 was about sentiment analysis on the Amazon Reviews Dataset. The dataset used for this assignment can be downloaded here.
The full problem statement can be found here.
Homework 2 was based on using HMM for part-of-speech tagging. The Wall Street Journal section of the Penn Treebank was used to build the HMM model for POS tagging. train, test and dev files were provided which are in the data folder.
The full problem statement can be found here.
In homework 3, the Amazon Reviews dataset was used to implement feed-forward neurals networks and recurrent neural networks using Word Embeddings. The dataset used for this assignment can be downloaded here.
The full problem statement can be found here.
The last homework was based on building deep learning models on named-entity-recognition (NER). The corpus and evaluation script were provided.
The full problem statement can be found here.