<![CDATA[Pharmaceutical company has become the most highlight company across the world lately because of the pandemic. Despite of the high demand market in pharmaceutical company, about 94% of large company across the world having difficulty in their supply chain that indirectly affect their services. The purpose of this research is to compare word embedding with compression model by doing sentiment analysis about the entity to find the best model that give better accuracy rates based on the opinion of Twitter, Instagram and Youtube, as they are the largest platform that its many users to express their opinions about an individual or an instance. Data is retrieved from Twitter, Instagram and Youtube using the R-Studio application by utilizing their API library, then preprocessing and stored in a database. Next step is labeling the data and then train the data using word2Vec and LSTM, GloVe and LSTM and lastly using Adaptive Compression (adaComp) to compress the both model word embedding. Unfortunately, we got imbalanced dataset after labeling process, so we add sampling technique to sampling the dataset using Random Under Sample (RUS) and Synthetic Minority Over-sampling Technique (SMOTE). After the data are trained and tested, the results will be evaluated using Confusion Matrix to get the best Accuracy. With several models that have been carried out,applying adaComp is proven to increase accuracy. In the Word2Vec word embedding with LSTM model, applying adaComp increasing its accuracy from 77% to 81%.]]>
Copyrights © 2022
