Sentiment Analysis of Union Budget 2019

Social media has become an effective tool in any organisation’s public reception. People are more engaged in the social life media nowadays thanks to availability of internet data. Online social media analytics is a powerful tool to boost e-commerce, politics etc. In marketing field companies use it to develop their strategies, to understand customers’ feelings towards products or brand, how people respond to their campaigns or product launches and why consumers don’t buy some products. Sentiment analysis also is used to monitor and analyse social phenomena, for the spotting of potentially dangerous situations and determining the general mood of the blogosphere. In political field, it is used to keep track of political view, to detect consistency and inconsistency between statements and actions at the government level. It can be used to predict election results as well! In this project we have a dataset of Union Budget 2019 of India that consists of tweets of public about Budget and label that opinion is positive or negative.

Disclaimer: Our sentiment analysis is purely based on collected data from tweets on the social networking platform “Twitter”. The data is collected on an unbiased method. This analysis is solely for educational purposes.

Scope:

In the following project, we use the sentiment140 dataset and uniondata dataset as training and testing data respectively. Firstly we perform the Lexicon-Based Sentiment Analyser to find out the count of positive and negative tweets from the union data.

Then comes the Machine Learning approach for sentiment analysis. In this technique, firstly the dataset are splited in X_train and X_test dataset to conduct training with X_train and test the performance with X_test. Following this, we clean up the test and training dataset and construct the feature vector of the training dataset. Then, we implement the classifiers constructing individual pipelines for each of them. Next, we train the dataset and store the performance metrics in a list. Lastly we show the positive and negative sentiment count for each classifier and see which classifier gives the best result.

Finally our aim is to answer our query, “The people like the new budget or not”.

Data Description:

Training Data Source:

Description: Sentiment140 Dataset (Last Accessed on: 27-07-2019)

Dataset name: train123.csv

Format: The data is a CSV with emoticons removed. Data file format has 6 fields:

0. The polarity of the tweet (0 = negative, 4 = positive)
1. The id of the tweet (2087)
2. The date of the tweet (Sat May 16 23:58:44 UTC 2009)
3. The query (lyx). If there is no query, then this value is NO_QUERY.
4. The user that tweeted (robotickilldozr)
5. The text of the tweet (Lyx is cool)

Test Data:

Description: This is a data collected specifically for this project.

Dataset name: test12.csv uniondata.csv tweetsdata.csv

Format:

1. 709 rows
2. 1 Column (tweets)

Requirements for data scraping:

• The Advanced Search URL for tweets on Twitter for Union Budget 2019

• Google Chrome with Web Scraper extension

• Twitter Advanced Search Scraper (Last Accessed on: 27-07-2019)

Code:

Filename: code.ipynb

Documentation

Filename: Sentimental Analysis of Union Budget 2019.pdf