Wikipedia is an online encyclopedia offering information on anything. It can be edited by anyone anywhere in fact that are over 185 edits per minute. This means that if there is a bias written into Wikipedia, this bias is also reflected in society as a whole.

""We loved working on this! Sentiment analysis!""
— Jerusalem,

To solve this problem we have created a web application called WikiDocter. We were inspired to do this because of the idea of sentimental analysis, which analyzes the positive, negative, and neutral words in a given situation. So to start off, you can first search anything that you would normally put into a Wikipedia search bar. Afterwards, two charts are generated, both comparing the positive and negative words in the intro paragraph of the article. The other feature we added was the save search bar, where once you press this, a new bar chart is created. You can put as many searches in this bar chart, allowing you to compare different topics and analyze which articles have more positive or negative bias.

Screen Shot 2015-08-10 at 10.23.38 AMFor example, with the 2016 presidential elections coming up, many people are turning to Wikipedia as an initial place to get information on candidates. However when we compared Hillary Clinton with Donald Trump we were astounded. While Hillary was mostly negative, Donald Trump barely had any negative words. When we added Bernie Sanders to the mix he stood as a middle ground between Hillary and Donald.

When so many people are trusting and using this website to gain awareness and are subconsciously inheriting this bias it becomes a really big issue. Wikipedia is designed to educate users so that they can do good and solve problems, but the inherent bias that is created is allowing for bigger problems to form. our web application informs people and showcases this bias before reading these pages they can visually see these charts and these biases can also be seen in our society as a whole.  This awareness helps to solve this problem.

Launch WikiDocter Now

Organization: Girls Who Code
Project Link: View Project Demo
Student: Jerusalem, Caiseen, Zoe
Project Team Members: Caiseen, Zoe, Jerusalem
Type of Work: Uncategorized
Location: Bay Area - Pixar
Grade: 12th Grade
Year Created: 2015

Girls Who Code aims to provide computer science education and exposure to 1 million young women by 2020. Make a donation today to help fund future programs.
Donate now »