niiiikou / NLP

This repository is about tokenization methods implimentation (n-gram, BPE, skipgram), sentiment analysis using RNN, LSTM, and GRU, using fairseq toolkit for machine translation and creating FAQ bot with Rasa.

NLP

This repository contains codes of my assignments that includes :

• Implementing n-gram language model and comparing white space, spacy and BPE tokenization methods.

• Developing skipgram model for generating semantic vectors.

• Sentiment analysis with neural networks (RNN, LSTM, GRU)

• Analysing the effect of changing ParsBERT architecture and parameters on NLI task.

• Sentiment analysis with zero-shot learning by using finetuned LabSE model

• Using fairseq toolkit for machine translation task (from english to persian )

• Creating FAQ bot on persian dataset by Rasa platform