in this project, I went through 3 Steps: 1-Listenning on tweets about Covid-19 in both languages arabic and french, and saving them in a csv file; 2-Stocking file in Azure Datalake; 3-Preprocessing and cleaning data for sentiment analysis
#1st Step: We Used Twitter Api for listening on topic that we want (Coronavirus, Corana, Covid-19) with condition searching in arabic and french language, taking tweets from accounts who have more than 500 followers and Saving them in csv file
#2nd Step: We programmed a timer who can upload the csv file to Azure Datalake every 5 min
#3rd Step Preprocessing by -Deleting stop words -Normalizing spaces -For arabic words removing "TACHKIL" -Removing hyperlinks -Removing Hashtags # and quote @ -Removing Retweets RT after this step we apply stemming
Finally we analyze sentiment using Deep Learning algorithms and Maching learning algorithms using scikit learn and tansorflow