Inspiration
I have always been an avid news reader. Once in a while a new topic comes up and I start to wonder what others think about it. This had led me to the creation of the Opinion Consolidator.
What it does
The Opinion Consolidator collects comments about the topic, then cleans the comments from punctuation and certain key stop words. The comments are then analyzed and the data is then put onto graphs and text for the user to be able to understand.
How we built it
It was built using Selenium Webdriver to obtain the comments. The comments are then cleaned from punctuation, emojis, and other stop words. The comments are then analyzed using TextBlob's polarity rating. The data is then put onto graphs made using the Matlibplot and put onto a GUI for the user's experience.
Challenges we ran into
One challenge that we ran into is the source of the comments. We had to find a way to reliably obtain comments about the topic and for there to be no limit for search. Platforms such as Twitter or Reddit had highly restrictive APIs that allowed for only a few thousand comments to be gathered per month. Using YouTube, we were able to access an unlimited amount of comments. However, YouTube doesn't have an API like Twitter or Reddit. This was averted by using Selenium Webdriver to find the most relevant videos about the topic and scraping the comments.
Accomplishments that we're proud of
I am proud of being able to use Natural Language Processing to analyze the data gathered and turn it into easy to read graphs and text.
What we learned
I learned about using Selenium's Webdriver. Using Selenium Webdriver allows you to input commands into chrome. This has high potential for possible actions, including web scraping, which the Opinion Consolidator uses.
What's next for Opinion Consolidator
The next step for Opinion Consolidator is to bring it onto a website for easier use.
Log in or sign up for Devpost to join the conversation.