Star

Inspiration

Given our experience with natural language processing, we wanted to see how that area could be applied to a business setting.

What it does

Star is a language moderation system that uses Google’s Perspective API to identify a post that is toxic on some way. The API gives rating on how toxic (vulgar or mean) the post was, whether or not it can be perceived as an attack on the author or another commenter, how likely the post is to be rejected and whether or not the post is spam, obscene, inflammatory, incoherent or unsubstantial. We also created an interface that displays the posts and shows if the API flagged the posts.

We believe that this system can be potentially very valuable for message boards of any kind, because it flags content that the administration might not want on there, such as threats of any kind or vulgar language.

How we built it

We built it using Google’s Perspective API and Django web framework.

Challenges we ran into

Learning how to use the Django framework to create the interface was our main challenge.

Accomplishments that we're proud of

We are proud of building this project, because it empowers administrators to easily monitoring and block unwanted content on their platforms.

What we learned

We learned how to use Django to create an interface, which is beneficial for us because it provides a streamlined process to create interfaces, thus saving us time in the future. We also learned how to use Google’s Perspective API, which is beneficial for us because it is a valuable tool for natural language processing that we might want to use in the future.

What's next for Star

We plan to make Star more adaptable. As administrators take action on specific posts, we want to rework that data back into Star's network so it can be smarter. We would like to reach out to administrators to see if they would like to use Star to moderate their content.

Built With

• django
• google-cloud
• google-perspective
• python