Skip to content

Shaon2221/Real-Time-Tweet-Analysis-on-US-Election-2020

Repository files navigation

Abstract

Social networking sites have become an indispensable part of the business and political campaigns. Therefore, it contains much information that can be used to predict market supply-demand or political popularity. People express their views on any issue on social media; those data, later on, can be utilized to understand the stance of the people on a particular issue. In this paper, we use real-time twitter data expressing views on issues of the USA Presidential Election. We explored the data to make the insight, and, then based on the insight, we applied the TextBlob library to find the polarity. What we have derived from the experiment, we utilize the outcome to interpret the data. Thus we reached an end commenting on both parties present situation. To complete this paper, we use the NLTK sentiment analysis package along with the TextBlob library of python.

Keywords

NLP, NLTK Corpus, TextBlob Sentiment Analysis, US Election 2020, Data Preprocessing

Objectives

  • Creating a report similar to conventional survey
  • Find out insights
  • Candidate popularity and engagement
  • Buzz and significant ideologies which creating impact
  • Opinion polarization
  • Visualizing real-time data plot
  • Finding which candidate is leading in the race.
  • Reasonable explanation
  • How it is potentially impacting on Election

Research Methodology

  1. Collect Data using Twitter API
  2. Clean Data
  3. Exploratory Data Analysis and Preprocessing

Sentiment Model Building


Results

Positive Percentage results respect to different party:


Democrats are just 3% ahead of Republican! 32% positve about Democrat whether 29% people positive about Republican. This election is neck-and-neck for Biden and Trump.

**Comparison between two major party in United States:**


It shows a large number of people may in cofusion about which party would be better! It is challenging for both party to convince them before election.

**Most frequent word used by people regarding US Election:**

From visualizing most frequent words from both parties, we can conclude that they are facing specifically same issues toward 2020 election. Black peoples and spreading hate may be critical for both. "Black Lives Matter" movement may play an important role. They are both using those words to fulfill their agenda.

Found more insights! Please, go through full project to explore them.

Tools Used

Python, Pandas, Matplotlib, Seaborn, NLTK, TextBlob

Any recommendation or correction is highly appreciated. Thank You

For Bengali Community:
আগামী নভেম্বর মাসে যুক্তরাষ্ট্রের প্রেসিডেন্ট নির্বাচন অনুষ্ঠিত হবে। নির্বাচন সামনে রেখে বিভিন্ন থিংক ট্যাংক এবং সংবাদ সংস্থা বিভিন্ন জরিপ পরিচালনা করে থাকে। হঠাৎ করে মাথায় আসে, এমন কিছু তো ডেটা সাইন্সের স্কিল ব্যবহার করে করা যায়। চেষ্টা করেছি, এরকম একটা রিপোর্ট তৈরি করতে। ডেটাসেট হিসেবে টুইটার এপিআই ব্যবহার করে ১১,০০০ টুইট সংগ্রহ করেছি। প্রজেক্ট সম্পূর্ণ করার মাঝে বেশ কিছু ইনসাইট পেয়েছি!
আমার এনালাইসিস অনুযায়ী, ডেমোক্রেট পার্টি সামান্য ব্যবধানে (৩%) এগিয়ে আছে। ডেটাসেট ডেমোক্রেটদের প্রথম কনভেনশন এর পর সংগ্রহ করা।এই প্রজেক্টের প্রগ্রামটা রান করলেই সেই সময়ের জন্য একটা রিপোর্ট তৈরি করা যাবে ৫-১০ মিনিটে। Isn't it awesome?🤔

ফিডব্যাক আশা করছি। 🙏
ধন্যবাদ 🖤

About

An End-to-end project of NLP

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published