train,valid=train_test_split(tweet,test_size= 0.15) Let’s generate some positive samples using the synonym replacement method.īefore data augmentation, we split the data into the train and validation set so that no samples in the validation set have been used for data augmentation. We can see that there is a small class imbalance here. Here, I will use the very same classification pipeline I used there but I will add data augmentation to see if it improves the model performance.įirst, let’s load the training dataset and check the target class distribution. In one of my previous posts, I used the data from this competition to try different non-contextual embedding methods. In this section, we will try data augmentation on Real or Not? NLP with Disaster Tweets competition hosted on Kaggle.

Keep in mind that data augmentation in NLP does not always help to improve model performance.Experiment to determine the optimal number of samples to be augmented to get the best results.A mix of different augmentation methods is also appreciated but don’t overdo it.Always try different augmentation approaches and check which works better.If you’re doing K-fold cross-validation, always keep the original sample and augmented sample in the same fold to avoid overfitting.Do not validate using the augmented data.NLPAug offers three types of augmentation: Let’s see how we can use this library to perform data augmentation. NLPAug helps you with augmenting NLP for your machine learning projects. In this section, I will introduce you to a python package that lets you do all these data augmentation easily and you can tune the level of augmentation you need using various arguments. Until now we have discussed many methods by which data augmentation can be used in NLP.īut effectively implementing these methods from scratch is a lot of work. You can check this wonderful notebook to see the complete implementation. There are many other transformations which you can try with this library. In this transformation, if the given text sample contains multiple sentences with duplicate sentences, these duplicate sentences are removed to create a new sample. In computer vision applications data augmentations are done almost everywhere to get larger training data and make the model generalize better. In this article, we’ll go through all the major data augmentation methods for NLP that you can use to increase the size of your textual dataset and improve your model performance. Augmentation methods are super popular in computer vision applications but they are just as powerful for NLP. What can you do to make your dataset larger?īut acquiring and labeling additional observations can be an expensive and time-consuming process.Īpply data augmentation to your text data.ĭata augmentation techniques are used to generate additional, synthetic data using the data you have. Experimental results show that the proposed model outperforms existing research on antonym synonym distinction in both speed and performance.There are many tasks in NLP from text classification to question answering but whatever you do the amount of data you have to train your model impacts the model performance heavily. Later, a classifier is trained based on features constructed from the distilled sub-spaces along with some word level features to distinguish antonyms from synonyms. We hypothesize that the pre-trained embeddings comprehend a blend of lexical-semantic information and we may distill the task-specific information using Distiller, a model proposed in this paper. We propose a novel approach entirely based on pre-trained embeddings. Existing solutions relying on large-scale corpora yield low performance because of huge contextual overlap of antonym and synonym pairs. National University of Defence Technologyĭistinguishing antonyms from synonyms is a key challenge for many NLP applications focused on the lexical-semantic relation extraction.

Sononym for classification