Inspiration

-Social media, such as Twitter, has many people expressing their opinions on a vast majority of topics ranging from love to Donald Trump. However, these tweeters follow no general trend from topic to topic. We wondered what exactly these patterns were after a team member of ours, Zach, suggested a program algorithm.

What It Does

-First, the user must type in a word or phrase. Our program searches for the last ten to the last few thousands tweets (depending on the range) that are related to our specified word/phrase. By utilizing a list of words that we know are inherently good and bad, we were able to assign values to unknown words to be used as inherently good or bad through association. Through continuous use of the program, the program becomes better at assigning the correct values to known words. This allows our program to score each tweet on its overall value of goodism or badism. This also contributes to the overall value of the goodism or badism the word or phrase has, as the program learns as it proceeds. The result Matrixcism values will then be posted online to the Twitter handle @ResultMatrix along with the most popular tweets for both goodism and badism at the Twitter handles @litmatrix or @dimmatrix, respectively. The AI will also analyze the level of emotion being used in each tweet and put that under an "Emotionism" value, which will be shown by the number of exclamation marks in the result matrix tweet.

How We Built It

-Through a combination of Python, C++, and the Twitter API.

Challenges we ran into

-Program being disorganized -AI initially blindly denoted combinations of good and bad as good overall or bad overall.

Accomplishments that we're proud of

-The successful execution of the program

What we learned

-We learned that Python is much more capable than we originally thought

What's next for Matrixcism

-Improve efficiency and implement a new UI for public (and private) usage in the future

Built With

Share this project:

Updates