It has been a while since our team wanted to implement an app that would ease the life of the muslim community and this ideathon pusshed us to start concretly.
What it does
Our solution provides basic features that we can find in a classic muslim app (mawaqit salat, qibla, quran...). In addition we added innovative features based on ia (chatbot for fatwas, ia moderator, Sound recognition of the 9ari, Tilawa correction, Classification of a7adith).
How we built it
Since it is a web app, we used html/css/Js/React for the front. We implemented the ia features with python and its many libraries/APIs (sklearn, tensorflow, NLTK, ...). For the back, we opted for the continuity on python, so we used Django with MongoDB as DBSM.
Challenges we ran into
Finding the right datasets for the work was really exciting to do since it is not very common to find a dataset of a7adith for example. Coding the chatbot was also challenging. Finally, we really enjoyed and learned from the feature extraction and preprocessing the a7adith for the classification.
Accomplishments that we're proud of
Our team is proud that we could find and implement really innovative features for the muslim community. We are also proud that our 'a7adith classification' was appreciated by our mentor and it pushed to go further on the implementation
What we learned
We got from this project some nice technical background in machine learning, deep learning, and web development. We also reinforced the communication between members and learned to share our idea to others while convincing them that it is a valid and reliable one.
What's next for Nebrasy
We will continue the realization of the project since it is on its early stage with the implementation of the remaining ai features. We will also transform the prototype of the app to a front web page and start our database. To do this, we will search for members from our beloved club (School of ai) that would be interested in our project.