/Sentiment-Analysis

Identified the sentiments of TripAdvisor reviews of a hotel

Primary LanguageJupyter Notebook

Sentiment-Analysis

Steps followed while performing Text Processing and Sentiment Analysis: a. Loading the necessary Libraries b. Loading the data c. Basic Text Processing (Case Conversions, Punctuation Removal) d. Tokenization e. Stop Word removal f. Top 10 high frequency words(food,good,service,buffet,staff,restaurant,breakfast,great,excellent) g. Calculated Polarity Score using Vader Sentiment Analysis h. Defined a function for Positive, Negative and Neutral Reviews(There are more positive reviews(88%) in the given data) i. Calculated percentage of polarity of reviews j. Identified most Positive,Negative and Neutral Review k. Vectorization l. Calculated TF-IDF(term frequency-inverse document frequency) m. Trained the model with all the data using Logistic Regression n. Model Evaluation using classification_report 0. Train Test Split.Fit on train data and predict on test data p. Creating a Data Pipeline and evaluating the model with classification_report