Merge pull request #156 from arhaanarif/main

Machine Learning/Zomato Restaurant Clustering and Sentiment Analysis

Machine_Learning/Zomato Restaurant Clustering and Sentiment Analysis/Readme.md

@@ -0,0 +1,46 @@
+# Title
+
+Zomato Restaurant Clustering and Sentiment Analysis
+
+## Description & Problem Statement
+
+This project entailed the utilization of advanced data analytics techniques to gain a deeper understanding of the restaurants and customer feedback on the popular online food delivery platform, Zomato.
+
+The problem statement for this project is to analyze and understand the restaurant industry in India by utilizing data from the Indian restaurant aggregator and food delivery start-up, Zomato. The project aims to gain insights into the sentiments of customer reviews, cluster Zomato restaurants into different segments, and analyze the data to make useful conclusions in the form of visualizations. The data analyzed includes information on cuisine, costing, and customer reviews. The project aims to assist customers in finding the best restaurant in their locality and aid the company in identifying areas for growth and improvement in the industry. Additionally, the project aims to use the data for sentiment analysis and identifying critics in the industry through the metadata of reviewers.
+
+## Goal
+
+The outcome of the analysis revealed that the restaurants within the city were grouped into five clusters based on their location, cuisines, and average cost for two. The sentiment analysis uncovered that, generally, customers held a positive sentiment towards the restaurants.
+
+## Conclusion
+
+Clustering and sentiment analysis were performed on a dataset of customer reviews for the food delivery service Zomato. The purpose of this analysis was to understand the customer's experience and gain insights about their feedback.
+
+The clustering technique was applied to group customers based on their review text, and it was found that the customers were grouped into two clusters: positive and negative. This provided a general understanding of customer satisfaction levels, with the positive cluster indicating the highest level of satisfaction and the negative cluster indicating the lowest level of satisfaction.
+
+Sentiment analysis was then applied to classify the review text as positive or negative. This provided a more detailed understanding of customer feedback and helped to identify specific areas where the service could be improved.
+
+Overall, this analysis provided valuable insights into the customer's experience with Zomato, and it could be used to guide future business decisions and improve the service. Additionally, by combining clustering and sentiment analysis techniques, a more comprehensive understanding of customer feedback was achieved.
+
+## Model
+
+I have chosen XG Boost model which is hyperparameter optimized
+
+## Evaluation Metrics
+
+- ROC AUC -0.818059
+- Precision -0.848111
+- Recall -0.894309
+- F1 -0.870598
+- Accuracy -0.835926
+
+## Libraries Required
+
+- Pandas
+- Numpy
+- Matplotlib
+- Sklearn
+- XGBoost
+- NLTK
+- Genism
+- Statsmodels