Computational Linguistics 1
All projects are implemented using Python 2.7
We implemented a word meaning disambiguation system based on 1) naive bayes and 2) peceptron. Given a word in a sentence and a set of possible meanings, the system predicts the most possible meaning for the word.
Packages used: ntlk, sklearn, numpy
We implemented a transition-based dependency parser (arc-standard). Given a sentence in the CoNLL format, it assigns the dependency relationship among words.
This project aims to provide hands on experience with neural models for NLP. In this project, we focus on transliteration from Bulgarian, which is written with the Cyrillic alphabet, into English.
Packages used: pytorch
These solutions are for reference only and should not be viewed by students currently enrolled in the course. If you read these solutions before completing the course, you will diminish the course's value for yourself and others.