rohitthapliyal2000 / amazon-mobile-sentiment-analysis

Opinion mining of Mobile reviews on Amazon platform

Amazon-Mobile-Sentiment-Analysis

Desciption :

This project works by scraping Amazon reviews for the user-desired mobile phone.

The first part includes searching the phone and scraping the reviews using infinite scrolling.

The second part includes training of NaiveBayesClassifier model on dataset and classifying the product reviews.

Tasks :

Searching the mobile phone on Amazon search bar

Finding the desired product i.e. 64GB, 128GB

Scraping the first 20(default, can be changed) pages of customer reviews sorted from new to old using infinite scrolling

Training the NaiveBayesClassifier algorithm on dataset

Classifying the reviews

Packages/tools used :

lxml library to use HTML element API

XML parser to parse the URL data into XML file

requests library to send HTTP request to Amazon webpage

NLTK library to use NaiveBayesAlgorithm and other features like bag-of-words, word-tokenizing

Requirements :

pip(Python Package Index) :

$ sudo apt-get install python3-pip

requests package :

$ pip3 install requests

lxml package :

$ sudo apt-get install libxml2-dev libxslt1-dev python-dev

$ pip3 install lxml

NLTK :

$ pip3 install nltk

To download nltk data, run following command on Python :

import nltk
nltk.download()

Select ALL and download from the pop-up GUI