NLP and machine learning has always inspired us. Though difficult, we decided to pursue a new challenge that could teach us something new.
What it does
The application converts the audio file in a video and a pdf to text in order to produce a summary of the contents. It is meant to be used as a starting point for students who want to revise a particular lecture.
How we built it
We focused on the mixed use of AI and openNLP, our goal was to recognize topics in text files and produce a meaningful summary. We used several libraries in order to build an easy-to-use tool that could be used by anyone.
Challenges we ran into
It was challenging to recognise topics and creating a framework to distinguish the meaning of a given paragraph. Although useful, openNLP does not solve all the problems that one could encounter when building such application. In order to make the program user-friendly, we decided to convert, cut and decode files in the backend, without the need for particular and somewhat frustrating format-converting tools.
What we learned
We found that the use of AI has helped us grow as computer scientists and students. It is a powerful device in today's software development and a flexible mean to achieve great results.