Bias Removal

Inspiration

My personal experience in the judicial system has been skewed by facts of how I act

What it does

It determines if the article is real or fake

How we built it

First I found a dataset to base it off of and decided which fields would be useful. Then once that was established a Natural language process was needed for the wording and help determining a pattern.

Challenges we ran into

Trying to use all of the fields turned out to be too much and didn't provide much use, so as a solution I broke it down into a vector problem and took only 2 fields.

Accomplishments that we're proud of

The percentage on the dataset was >90% which rarely happens on my algorithms.

What we learned

I didn't know vectorization until I tried it on this project as it is my first NLP project as well.

What's next for Bias Removal

For sure there were more fields that we could have used to make it more accurate, and the dataset provided was very small so a bigger dataset would leave room for Supervised learning.

Built With

• javascript
• python