VIS-NLP: Vaccination Inventory System for justified user using Natural Language Processing

Abstract: In the healthcare industry, especially the Covid-19 pandemic in 2020, produced huge problems with isolate patient and patient heath. Thus, created large amount of data that has been generated every day for the patient heath, in this case is to justify the vaccination of users from social network Twitter. Processing such large volume of the data involves high computation overhead. Good health and well-being; to ensure healthy lives and promote well-being for all at all ages is United Nations 3rd Sustainable Development Goal and we want to align our study with it as well. It is crucial to create an application that is beneficial for humanity health. When we get large datasets from pandemics like Covid-19, for large scale datasets, we presented a solution to verify the user if they are vaccinated or not vaccinated by using Natural Language Processing methods to build an accuracy result, we tried to reduce the computation overhead by storing the data in distributed environment. After processing data, training the data, used pad_sequences, Keras, NLP to build the model. Through multiple epochs we have got an accuracy towards 90 to 91% (which is closer to state-of-the-art methods i.e., 95%). And since our accuracy is higher, we can further utilize it to increase for higher number of epochs. We hope scientists can further develop it and use it in real world applications so that more precious human being lives can be saved. By implementation of its successful results, it also aligns with one of the United Nations Sustainable Development Goals i.e., 3rd: Good Health and Well-Being.

Keywords: Natural Language Processing, Keras, Tokenizer, UNSDG

DOI: 10.54941/ahfe1003459

Cite this paper:

​