Data Science


Software Development 

Sentimental Analysis

Analysis of personal profiles to derive data, specifically from the tweets posted by the user, add token values to each data. It uses the Natural Language Processing technology to extract, quantify and study the data. It is later interpreted from the user’s post and likelihood.

Client Profile

Our client is a leading software developing company engaged in building its enterprise solutions and risk management in the most efficient way. They are committed to grow, become profitable and socially responsible. They rely on innovation, responsiveness and quality as the key to success.


The vital role of the project is to reveal the user’s nature by using the historical data of tweets that has been posted. The nature of the user will determine whether the user has used negativity or positivity in the major part of their tweets.


Key Challenges

The major challenge faced was in deriving the value of a user with the tweets he posts which is screened by the semantic analysis process. The preprocessing of the data via API in twitter was a very intricate task where the removal of stop words was needed. Identification of the tweet was another crucial task where we had to recognize each user in it.

Our Approach

We used the NLTK sentiment analysis package which contains the steps for processing functions. The “kwargs” were termed as one of the additional parameters for that function. The bigram combines its return with the simple bigram frequency from the user tweets. The extraction of the individual tokens is stored as key points at the libraries in the form of unigrams. Words are stored in the form of Bag of words which are consequently used for determining the nature of the user with their tweets.

The Solution

We analyze the sentiment of a user by determining to whether add it in the context through twitter ID of the user. If a positive result is observed, it recommends that the user has tweeted more in favor of positive words and it is vice versa if a negative trait is observed.


An efficient sentiment analyzer is developed by adding credentials of the user into any scoring system. Additionally, it follows a developed approach which reduces space and time complexities which results with a high accuracy in determining the user’s credentials with tokens from the tweets.

Technology Used

  • Back-End: Python - NLTK
  • API: Twitter API
  • Front-End: Django sc


The twitter data is derived from the API which is later passed into the sentiment analyzer which instead verifies whether a user can be added or not into the process. If a user holds a positive note to his twitter id it increases the return score leading it to hold a positive score and a negative note implies the return score to become negative with a zero or a lower score than the origin.

Download Sentimental Analysis Case Study


Send download link to:

Other Case Studies