Drug reviews provides valuable information like the side effects of the drugs, how that particular drug helped someone cure the disease or whether that drug was helpful or not, etc. Based on the past experience or past reviews given by the people who took the drug, we can help classify which reviews are positive and negative using sentimental analysis techniques on the Drug review Dataset. This Dataset includes drug names, conditions for which the drug is used, patient reviews for that particular drug, rating of the drug, the date of the review and also the number of User counts who found that drug useful. Before taking any drug, this might help a person to understand whether that particular drug has a positive or negative side effects and can relate the same situation with the past experience of that person. We will be using different libraries and approaches and will be comparing this approaches which might help us gain information about how that particular library is stemming or lemmatizing a particular word to reduce its inflectional forms into a common base form like ‘colors’ into ‘color’, ‘different’ into ‘differ’ and also whether that particular word has a positive, negative or neutral meaning. Also, using different classification models in our approach will help us understand which model is performing better.
Link to the dataset https://archive.ics.uci.edu/ml/datasets/Drug+Review+Dataset+%28Drugs.com%29