Inspiration
Our inspiration is based on motivate end user to help us to collect label data, and use those data to train a learning model to help the officer to quickly identify who is the pimp and then find the victim through the pimps.
What it does
It helps to quickly identify the pimp's post
How we built it
We build an app to link those posters and let the user label them. After getting the labeled data, we use NLP tools to pre-process the input data and convert it to the vectors. Using those data with the labels provided by the users, we are able to train a random forest classifier.
Challenges we ran into
In the beginning, we have to label data by hand. The input data are all HTML files and we use python to clean the data and extract only the useful part from it.
Accomplishments that we're proud of
We success to build an app implement the functions that we want and gain a random forest classifier.
What we learned
The team work, and processing data techniques, using NLP tools to handle words data.
What's next for human trafficker's post identify
We try to improve the randomforest classifier, currently due to lack of training data our model suffer from the overfitting. We hope that, we are able to gain more labeled data and apply more complicate algorithm such as deep learning to achieve a higher accuracy.
Log in or sign up for Devpost to join the conversation.