Inspiration
My main inspiration is my religion prof (elective class) that keeps taking off marks for having bad titles. So we decided to make a program that will evaluate how good the title is.
What it does
It gives a grade from 1 to 10 based on how likely your title will make the paper successful.
How we built it
Using a dataset of over 170 000 articles, we used their performance score and their titles in order to train an AI model so that it can find a pattern between the title and good/bad performance of the article. Then we need to classify it and give it a number from 1-10
Challenges we ran into
We have no experience with training a model. Also, we tried using Jupyter and it kept crashing on us, so we lost our progress numerous times. We changed platforms multiple times, but none of them seemed ideal. Our model doesn't output one number, but an array.
Accomplishments that we're proud of
We trained a model for the first time. We learned about multiple ai libraries; we got a pretty good introduction to artificial intelligence model trainings.
What's next for Is This Title Good?
Use more dataset to train the model in order to achieve better accuracy, use a more diverse dataset and polish the model in order for it to output a singular number.
Log in or sign up for Devpost to join the conversation.