About Twitter Sentiment Analysis
The client has a political background, works as a public figure and has a large number of followers on social media.
Services Provide
- Web Application
- Web Scrapping
- Natural Language Processing
Technologies Use
- Python
- Scikit-learn
- Django
- MSSQL Server
Business Problem
Understand people sentiments on social media after launching different political campaigns.
Challenges
- User doesn't want to log in via twitter account so cannot access Twitter API to fetch tweets
- Tweets have different grammatical constructs and sometimes may have non-english words written using english characters
- Twitter has certain security measures which blocks the scrapping bots
- Build and adjust machine learning algorithm as it processes more tweets
Solutions
- Extraction - Fetch tweets using various web scraping libraries Scrappy, BeautifulSoup and Selenium.
- Preparation - Prepare dataframe for preprocessing the tweets to remove non-contextual words
- Exploration - Identify the most tweeted hashtags and most used words, Find Collocations (Words most frequently used together).
- Clustering - Group the tweets together by distance based on the words used.
- Parsing the tweet messages processed and generate the training dataset using gensim.
- Processing the content further and fetch the topics (Group of words) using sklearn.
- Identifying and categorizing in different sentiments using TextBlob and nltk.
Need a similar app? But first, get your FREE QUOTE...
Recent Blogs
Stay updated with our latest blogs on Python.
