Text Extraction In Social Media
BY:-RAVINDRA CHAUDHARY
SACHIN SINGH
UNDER THE GUIDENCE OFMRS. SMITA TIWARI
CONTENT
Introduction Problem Statement Objective Methodology Module Future work and scope Conclusion
INTRODUCTION
What is Sentiment Analysis…??
It is the classification of the polarity of given text in the document. The goal is to determine whether the expressed opinion in the text is
Positive,Negative or Neutral.
For Example:- Positive :- sarvjeet is good guy… negative :- jasleen is misusing the law.. Neutral :- waiting for court decision..
Problem Statement Sentiment analysis is classifying the polarity of given text in a document in a
sentence is positive ,negative or neutral.
To collecting data and categorizing into different sets for different purposes.
Making sentiment tool for measuring all sentiments by one tool.
Increase the accuracy of the result which is measured by sentiment tool.
Objective
To implement naïve baye’s Algorithm for classification to text polarity into Positive , Negative,or Neutral sentiments.
Using different type of NLTK classifiers for classifying with more accuracy.
Methodology1. DATA COLLECTION download the tweets using Twitter API.
2. TOKENSIER Twitter using POS(part of speech) tagger.
3. PRE-PROCESSING Remove slag(non-english) words Replacing emoticons by their polarity. Remove URL and HASTAG(#),numbers. Peplace sequence of repeted character coooooool by cool. Remove noun and prepositions.
FEATURE EXTRACTION Percentage of capitalized word No of –ve/+ve capatilized word No of +ve/-ve hastag No of +ve/-ve emoticons No. of negations No. of special characters ex..@#%^*
CLASSIFICATION AND PREDECTIONS The model is built to predict the sentiment of new tweets… Feature extracted are next focused to classifier
MODULES OF PROJECT
Module 1:- Extracting data from social media. Module 2:- Tokensing fetched data. Module 3:- Preprocessing fetched data. Module 4:-: feature extraction of data. Module5:- Classification of data.
MODULE 1:- We are using facepager to extract the data from social media. The following process shows how to fetch data from social media
Future work and scope
Web application can be converted to mobile applications Sentiment analysis may be implemented in futyre for accuracy
purposes Updating dictionary for new synonyms and antonyms Data pre-processing using more parameters to get best sentiments
Conclusion
We conclude that using different classifiers it is easier to classify the tweets and documents.
By improving the data sets we get more accurate results (sentiments).
THANKYOU EVERYONEFor taking interest in my slides