Token and Sentence Level Classification with Google's BERT (TensorFlow)

BERT-Text-Features for Tokenized Transcripts from P2FA.

Exploited GoogleBERT embeddings on LIAR dataset for multi-class classification task of Fake news detection.

Implementing Multilingual WSD using [Normal, Atten]BiLSTM, Seq2Seq[Atten], Multitask WSD

This project uses BERT(Bidirectional Encoder Representations from Transformers) for Yelp-5 fine-grained sentiment analysis. It also explores various custom loss functions for regression based approaches of fine-grained sentiment analysis.

A showcase of combining Elasticsearch with BERT on the HackerNews public data

Low resource machine translation using Transformers and Iterative Back translation

Problem Statement: Given the tweets from customers about various tech firms who manufacture and sell mobiles, computers, laptops, etc, the task is to identify if the tweets have a negative sentiment towards such companies or products.

This project is submitted as python implementation in the contest of Analytics Vidhya called "Identify the Sentiments". I enjoyed the joining of this competition and all its process. This submited solution got the rank 118 in the public leaderboard.

Fake News Pair detection challenge on kaggle

A very simple framework for state-of-the-art Natural Language Processing (NLP)

Chinese Word Segmentation task based on BERT and implemented in Pytorch

Python client to use SOCO answer-as-as-service platform.

Source code for "Impact of Spanish Dialect in Deep Learning Next Sentence Predictors", CLEI Panama, Duboue (2019)

BERT-based thesaurus.

Code for the paper "How Relevant Are Selectional Preferences for Transformer-based Language Models?" (Metheniti et al., 2020)

This project uses Support Vector Machine and BERT word embeddings to make sentiment prediction on messages from stockttwits, one of the biggest financial forum in the world.

COVID-19 Question Dataset from the paper "What Are People Asking About COVID-19? A Question Classification Dataset"

Use ELMo-like Bert word embeddings to match a word's contextual use to the appropriate Merriam-Webster definition

This repository is the PyTorch implementation of the Attention-Enhanced Relational Graph Convolutional Networks method for the task Multi-lingual and Cross-lingual Word-in-Context Disambiguation from SemEval-2021.