Inspiration
I always thought that people will require a machine learning solution which parses the text as toxic or non toxic. Most of the parents are worried about their childrens online presence. i.e. is someone will bully their child on social media platforms like facebook or twitter. OR may be the child reads something that he/she should not read. To overcome this issue I have developed this API. Though I was seeing people building apis and applications using Python. But I always wanted to implement these things in java.
What it does
It parses the given sentences ito either toxic/negative/sensitive(0) or non toxic(1) in microseconds.
How I built it
I built it using quarkus and apache open-nlp library and by providing the dataset of toxic and non toxic sentences. and http://che.openshift.io/ used this to write all the source code. I found that quarkus already has presence over there so that was a bit easier to move ahead.
Challenges I ran into
At the beginning I was not aware about how I can use open-nlp to implement this thing. But when further checked , I found some syntax and relevant classes.Also I was not aware about maven and quarkus. But by going through quarkus tutorials I was able to make it.
Accomplishments that I'm proud of
I completed the machine learning solution for sentiment detection using java.
What I learned
I learned Quarus,Maven, eclipse che, Also I learned how to host this application on cloud .
What's next for Sentiment-detector
We will look forward to implement emotions analysis also in this.
Log in or sign up for Devpost to join the conversation.