Muhammad Saad Amin's repositories
Urdu_Meaning_Bank
A very first semantic resource for Urdu.
augmentation-for-semantic-parsing
Augmentation for Semantic Parsing on PMB-5.0.0
Test-set-duplication-sentence-checking
This code finds the duplicate sentences while tense and name augmentation in actual test file to avoid evaluating biased test set.
Tense-Verb-Augmentation-for-DRS-and-Text
This code perform tense-change i.e., verb augmentation in DRS and Text files separately.
gMLP
A code to implement gMLP on ASL data.
Bert-Score-for-neural-translation
This code calculates the Bert-Score for DRS-to-Text based model generated sentences.
Name-Augmentation-for-Proper-Nouns-or-Named-entities-from-DRS-and-Text
The code performs Name Augmentation for Proper Nouns by replacing Person, City, State, and Country Names from dataset with most popular names in the world.
SST-and-Non-SST-based-different-Strategies-to-Extract-Nouns-in-Gold-PMB-dataset
This code have different strategies to extract Nouns from dataset. It also contains code to get Super Sense based approach to get Nouns with different contextual sense. It also have the code to get BERT based contextual replacement of Nouns.
Nouns-present-in-multiple-contextual-catagories-of-Tags-through-SST
This code will deal with the Nouns having multiple senses. A special case of Noun Unlexicalization where same Noun can have different Super Sense based Categorical Tags.
Calculate-Sentence-by-Sentence-BLEU
This code calculates sentence by sentence BLEU scores for DRS-to-Text generation Task.
Tweeter-based-Flood-or-Earthquake-Sentiment-and-Emotion-Classification
The code scrap data from Tweeter and perform Binary and Multi-sentiment and also perform emotion classification.
k-fold-cross-validation-in-Unlexicalization-of-DRS-to-Text-Generation
This file contains all the codes to perform K-Fold Cross Validation on Noun's Unlexicalization experiments.
Unlexicalizing-and-relexicalizing-Nouns-in-DRS-to-Text-Generation
Extracting Named Entities and Common Nouns from dataset in perform unlexicalization. And also substituting back the lexical information in model generated unlexical outputs.
Gold-PMB-Augmented-Dataset
This repository have augmented dataset files for all formats of augmentation applied on Gold-PMB dataset.
pre-processed-gold-pmb
this is pre-processed gold pmb dataset to conduct baseline experiment for DRS Data Augmentation
NER_CoNLL
Named Entity Recognition on CoNLL 2003 dataset
gdelt
Dataset generation from GDELT using BigQuery
GMB_2_NER
implementation of NER on GMB-2.2.0 dataset using LSTM, bi-LSTM, LSTM-CRF