Creepdetect

Inspiration

Sometimes (most of the times), girls receive really creepy messages from guys on the dating app, Tinder. We wanted to troll some of them back

What it does

The algorithm consists of a machine learning way to detect if messages received from matches pass a threshold of creep level or not. This machine learning method was built on a database of bad pick-up lines. if the message passes a creep threshold, it send a Drake lyric back to the match

How we built it

We used a Tinder package from Github and modified it to only send messages back to matches who sent messages to me. We built the machine learning method using sci kit learn

Challenges we ran into

Integrating with the Tinder API to only run the algorithm when a match sends me a message first

Accomplishments that we're proud of

That we were able to interact successfully with the API and extract the matches sent to me

What we learned

develooped python skills and played around mobile and web apps

What's next for Creepdetect

maybe personalize the response sent back to creepy messages instead of an arbitrary message

Built With

• api
• python
• python-package-index
• tinder