Project Title

Amazon-Reviews Mobile phone Reviews

         Fig: 1 Shows different Sentiment 

Motivation

Why some brands are so popular and why some brands are not doing well based, on review trying to understand why some popular brands makes the difference and some other brands fall back.

Sentiment Analysis of Mobile Phones reviews Using NLTK

Using this sentiment analysis I want to show that why some major brands are leading and some brands are not doing well. Based on review people try to convey a message that some brands are lacking based on their performance, feature as well as quality.

Installations

Importing libraries

Python

Jupyter notebook (label: good first issue)

Jupyter interactive notebook

Pandas (label: good first issue) Flexible and powerful data analysis / manipulation library for Python, providing labeled data structures similar to R data. Frame objects, statistical functions, and much more.

numpy (label: good first issue)

pip install numpy

It is the core library for scientific computing, which contains a powerful n-dimensional array object.

Scikit-learn is a machine learning library for Python.

pip install scikit-learn

NLTK(Natural Language Toolkit) The Natural Language Toolkit, or more commonly NLTK, is a suite of libraries and programs for symbolic and statistical natural language processing (NLP) for English written in the Python programming language.

pip install nltk

Running the tests

Reading csv File. Since File was available on kaggle. There are 4410 phone models in this data set. There are 385 brands in this data set.

    Fig: 2 Different types of Phone Company in the list 

Cleaning data by removing rows having null values

General Description of data

  Fig: 3 Shows Min max Description 

Top 10 brands in the data set sorted on the basis of sum of Ratings.

  Fig: 4 Shows Top Brands 

Correlation between price & rating

 Fig: 5 Shows different Price and Rating

Correlation between Price and Review Votes

  Fig: 6 Shows different Price and Review 

Correlation between Rating and Review Votes

  Fig: 7 Shows different Rating and Review

It is observed that Rating has a NEGATIVE CORRELATION with Review Votes = -0.046526

   Fig: 8 Shows different Rating 

It is observed that Rating has a POSITIVE CORRELATION with Price = 0.073948

   Fig: 9 Shows different Rating of Positive

   Fig: 10 Shows different item  

NLTK function to find sentiment value and sentiment

   Fig: 11 Shows different Top brands Reviews 

  Fig: 12 Shows different Sentiments 

   Fig: 13 Shows different Accuracy 

   Fig: 14 Shows different intensity 

Observation: Sentiment variation is concentrated towards positivity

   Fig: 15 Shows different sentiment based on positivity 

   Fig: 16 Shows Bestselling Brands 

   Fig: 17 Shows different product name and sentiment value 

  Fig: 18 Shows different Values 

Sentiment Analysis for Top 5 brands

  Fig: 19 Shows different Sentiment  

Observation :

1. Sentiment concentration towards positivity decreases as we move from top to lower brands.
2. Population towards negativity and neutrality keeps on increasing as we move downwards.¶