-
-
Main Page
-
Results Page
-
Emotions
-
Wordclouds
Inspiration
Data scientists sometimes need latest specific data to understand patterns and come to a conclusion. But many datasets on many topics are very very old. So, I wanted to build a tool that can collect and provide them with latest and time specific data along with analysis of that data.
What it does
It asks for a topic , timeline, number of data points and then fetches the data from twitter based on the entered parameters. The data can be dowloaded in CSV , PDF and JSON format. The next big feature is data sentiment anlysis . The complete visual analysis is done as well using matplotlib but later it will be shifted to plotly for interactive plotting. It will help people to understand what are the views of a large number of population on a specific topic.
How I built it
I build it using Javascript, HTML , CSS for frontend and Python and Flask for backend.
Challenges I ran into
I build this project in the given timeframe only. So, many features are still incomplete. But the complete analysis is done in the jupyter notebook using sample twitter data. Just integration with the frontend for analysis page is left. It will be done using Plotly and Cufflinks.
Accomplishments that I'm proud of
I have built the complete application alone. I have successfully deployed the application with continuous deployment on heroku. The complete analysis is done in the Jupyter Notebook. I have added the .tech custom domain as well. I will surely work on this applciation and add more and more updates to it.
What I learned
After research I learned that this is a need for many data scientists out there. I learned on how to connect domain.com to heroku. I learned that plotly and cufflinks are used for interactive plotting. I learned a lot about sentiment analysisas well.
What's next for Tweetify
Adding more features like donwloading the generated/fetched data in multiple formats. Generating an interactive panel for user to interact with the plots that are generated. Adding sentiments column to the generated data.
Log in or sign up for Devpost to join the conversation.