Facebook Data Analysis Using Hadoop - Facebook Data Analysis Using Hadoop Project

Facebook Data Analysis Using Hadoop Project

Posted on

Fb Information Evaluation Utilizing Hadoop


To investigate the Fb information utilizing Hadoop for the aim of higher resolution making within the enterprise.

Challenge Overview

Sensible telephone with out social media utilization in every day way of life of individuals is unthinkable. That a lot impact has been created within the way of life of individuals by smartphone and social media. There are various social media comparable to Fb, Twitter, and so on., As per 2017 statistics, practically 1.37 billion every day lively customers for Fb. Each consumer contributes some sort of information to in structured or semi-structured or unstructured information format. Enterprise  homeowners make the most of this information to know buyer want and their conduct to make revenue of their enterprise. Fb information evaluation is the method of gathering, analyzing Fb information and visualizing extracted outcomes to the top consumer.

The consumer information is collected from Fb primarily based on their actions. Person conduct, variety of likes, variety of posts, sort of posts, their feedback, and so on. are saved by the database server. Feedback by the consumer in unstructured codecs, whereas different information in structured and semi-structured format. Petabytes of information is generated by Fb customers. So Hadoop, MapReduce and associated massive information ideas used on this project to investigate the information.

Proposed System

The proposed system focuses on analyzing sentiments of Fb customers utilizing Hadoop. The consumer sentiments collected are categorized into constructive, detrimental, impartial.The proposed system structure is proven within the determine.

Facebook Data Analysis Using Hadoop - Facebook Data Analysis Using Hadoop Project

                                                Determine: Proposed System Structure


Module 1: Fb API

Fb API is used as an authentication API to extract the consumer contents associated to the question requested.

Module 2: Information Pre-Processing

Information Assortment: The information are collected from Fb utilizing Hadoop by way of the Fb API primarily based on the requested question.

Information Preparation: The collected information consists of various feelings, cease phrases, acronyms, and so on. However throughout evaluation the sort of information must be transformed into the right format to extract sentiments from the consumer conduct.

  • Tokenization
  • Varied Dictionaries
    • Acronym Dictionary
    • Cease Phrases Dictionary

  • Emoticon

Think about one of many Fb posts relating to new cellular options. Customers opinion in regards to the new telephone may be constructive or detrimental or impartial.

Instance for Optimistic Sentiment

Seems are superior.Battery backup is superb. Digicam is sweet.The show mild high quality is sweet.

Instance for Impartial Sentiment

Though that is good cellular, seems good, however Downside is that it doesn’t present separate House for twin SIM & reminiscence card collectively.

Instance for Damaging Sentiment

Not good one as anticipated. Digicam high quality very poor.


Feedback extracted from Fb are divided into tokens. This is named tokenization course of. For instance, ‘Seems are superior. Battery backup is superb. Digicam is sweet. The show mild high quality is sweet.’is split down into ‘Seems’, ‘are’, ‘superior’, ‘.’, ‘Battery’, ‘backup’, ‘is’, ‘wonderful’, ‘.’, ‘Digicam’, ‘is’, ‘good’, ‘.’, ‘The’, ‘show’, ‘mild’, ‘high quality’, ‘is’, ‘good’, ‘.’

 Acronym Dictionary: It’s used to offer the required acronym for the phrases, if wanted.

Cease Phrases Dictionary: It’s used to take away the unrelated phrases within the sentiment evaluation course of. Instance: A, An, The, Has, Are, Is.

Emoticon:That is used to detect the emoticons for the aim of classifying the remark as constructive or detrimental or impartial.

 Module 3: Sentiment Evaluation

The consumer sentiments collected from the Fb are categorized into constructive, detrimental, impartial. This sentiment evaluation will be carried out for various functions primarily based on the enterprise aims.

Module 4: Information Visualization

After the Fb sentiment evaluation, the extracted and analyzed sentiments are visualized utilizing Tableau.

Software program Necessities

  • Linux OS
  • Hadoop & MapReduce
  • Fb API
  • HIVE
  • Tableau

{Hardware} Necessities

  • Exhausting Disk – 1 TB or Above
  • RAM required – 4 GB or Above
  • Processor – Core i3 or Above

Know-how Used

  • Large Information – Hadoop

Supply projectgeek.com