The proposed work has retrieved tweets on a particular disease and treatment combination from twitter and they were processed to extract the sentiments. Initially the polarity values were set up in a range from weakly negative to strongly positive and the tweets were analyzed. The overall sentiment of the tweets related to breast cancer and chemotherapy was weakly positive. Naive Bayes algorithm was applied on the tweets retrieved on the same disease and treatment combination. Nearly 10,000 tweets were analyzed using the Pubmed and Google book search engine as a training corpus. The sentiments were plotted in the graph which shows that the sentiments were neutral. Lastly, to find the most occurred word in the tweets, bigrams were used and cooccurrence of words were plotted using Natural Language Tool Kit in python.
展开▼
