This document discusses using machine learning and deep learning techniques for sentiment analysis on social media data. It proposes a system to classify tweets as having positive, negative or neutral sentiment. The system involves data acquisition from Twitter, preprocessing tweets, extracting features, and applying machine learning classifiers like SVR, Random Forest and Decision Tree. It aims to analyze public sentiment on topics and help organizations understand people's views.

1. Venkat Java Projects
Mobile:+91 9966499110 Visit:www.venkatjavaprojects.com
Email:venkatjavaprojects@gmail.com
Sentiment Analysis using Machine Learning and Deep Learning
Abstract:
With the increasing rate at which data is created by internet users on various platforms, it
becomes necessary to analyze and make use of the data by the Defense and other Government
Organizations and know the sentiment of the people. This shall help the organizations take
control of their actions and decide the steps to be taken shortly. Added to it, when something
crucial is happening in the nation, it is of paramount importance to decide every step without
hurting/violating the sentiments of the people. In the era of Microblogging, which has become
quite a popular tool of communication, millions of users share their views and opinions on
various day-to-day life issues concerning them directly or indirectly through social media
platforms like Twitter, Reddit, Tumblr, Facebook. Data from these sites can be efficiently used
for marketing or social studies. In this paper, we have taken into account various methods to
perform sentiment analysis. Sentiment Analysis has been performed by using Machine Learning
Classifiers. Polarity-based sentiment analysis, and Deep Learning Models are used to classify
user's tweets as having `positive' or `negative' sentiment. The idea behind taking in various
model architectures was to account for the variance in the opinions and thoughts existing on such
social media platforms. These classification models can further be implemented to classify live
tweets on twitter on any topic.
Existing System:
 In recent years, researchers preferably made the use of social data for the sentiment
analysis of people's opinions on a product, topic, or event. Sentiment analysis, also
known as opinion mining, is an important natural language processing task. This process
determines the sentiment orientation of a text as positive, negative, or neutral.
 Twitter sentiment analysis is currently a popular topic for research. Such analysis is
useful because it gathers and classies public opinion by analyzing big social data.
However,Twitter data have certain characteristics that cause difculty in conducting
sentiment analysis in contrast to analyzing other types of data.
 Tweets are restricted to 140 characters, written in informal English, contain irregular
expressions, and contain several abbreviations and slang words. To address these
problems, researchers have conducted studies focusing on sentiment analysis of tweets .
 Twitter sentiment analysis approaches can be generally categorized into two main
approaches, the machine learning approach, and a lexicon-based approach. In this study,
we use machine learning techniques to tackle twitter sentiment analysis.
Proposed System:

2. Venkat Java Projects
Mobile:+91 9966499110 Visit:www.venkatjavaprojects.com
Email:venkatjavaprojects@gmail.com
we present the methodology used in this study. The proposed system is basically composed of
four main modules. The _rst module is data acquisition, which is a process of gathering labeled
tweets to perform sentiment analysis; the second module, this dataset undergoes various steps of
preprocessing to transform and re_ne tweets into a data set that can be easily used for subsequent
analysis. The third module concerns the extraction of relevant features for building a
classi_cation model. Then, the balancing and scoring tweets technique is illustrated. The last
module is applying different machine learning classi_ers that classify the tweets into high
positive, moderate positive, neutral, moderate negative, and high negative. Figure 1 shows the
various steps performed for sentiment analysis using machine learning algorithms.
System Architecture:
Modules:
1. Load NLTK Tweets
2. ReadNLTK Tweets
3. Run SVR Algorithm
4. Run Random Forestand Decisiontree

3. Venkat Java Projects
Mobile:+91 9966499110 Visit:www.venkatjavaprojects.com
Email:venkatjavaprojects@gmail.com
5. DetectSentiment Type
1) Load NLTK Tweets: Using this module we will load twitter sentiment
corpora dataset from NLTK library.
2) Read NLTK Tweets: Using this module we will read tweets from NLTK and
then clean tweets by removing special symbols, stop words and then perform
stemming (stemming means removing ing or tion from words for example
ORGANIZATION word will become ORGANIZE after applying stem) on
each words. Then we will calculate TFIDF vector.
3) Run SVR Algorithm: In this module we will give TFIDF vector as input to
train SVR algorithm. This algorithm will take 80% vector for train and 20%
vector as test. Then algorithm applied 80% trained model on 20% test data
to calculate prediction accuracy.
4) Similarly we will build model for Random Forest and Decision tree to
calculate their accuracy.
5) Detect Sentiment Type: Using this module we will upload test tweets and
then application will apply train model on those test tweets to predict
sentiment of that tweet.
SYSTEM CONFIGURATION:
Hardware requirements:
Processer : Any Update Processer
Ram : Min 4 GB
Hard Disk : Min 100 GB
Software requirements:
Operating System : Windows family
Technology : Python 3.6
IDE : PyCharm