Data analysis

Part I: Given Twitter Dataset (extracted from Kaggle) • Create a data frame and call it Temp_Sent and load all data from the given Twitter Dataset (all 1.4 million twits) • Make shuffling (changing the orders of the rows) of Temp_Sent (use the following statement Temp_Sent = Temp_Sent .sample(frac = 1) • Create a data frame and call it Sent and any copy 100 k twits from Temp_Sent • Create a function to Make text cleansing of the text column • Use CountVectorizer() and MultinomialNB() to create sentiment prediction model and do the required test • Use TfidfVectorizer () and MultinomialNB() to create sentiment prediction model and do the required test • Use confusion_matrix to evaluate both models Part 2: PR Dataset (found in black board) PR is a dataset about the evaluation of some models of smart phones • In the product column, show only the product general name: for example :Samsung, hawawi, Iphone and remove other text • Show in a bar chart, the average score of each model • What is the best model in each country