Developed a AI machine learning algorithm that accurately predicts user
preferences based on input queries, offering a list of top movie
recommendations.
Data Utilization:
The project utilizes two TMDB dataset files: tmdb_5000_credits and
tmdb_5000_movies, incorporating information of movie attributes such
as include cast, keywords, director, and genres, facilitating a
comprehensive understanding of movie characteristics.
Algorithmic Approach:
Data preprocessing involves merging datasets based on ID columns and
extracting relevant features to construct a comprehensive data
frame.
Count Vectorization is implemented to quantify word frequency
distribution in movie descriptions. Cosine similarity of the feature
matrix is calculated, providing a measure of similarity between
movies.
The system's database is limited to 5000
movies, restricting its ability to recommend movies
outside this dataset.
eMarket
Web Application
Language Used:
Frontend: JavaScript, jQuery, Razor, and CSS
Backend: ASP. Net Core, C#, and MS SQL
Led a team of 4 to design and develop a Web application using Visual Studio. e-Market is a web application that is an eCommerce platform where users can post classified ads to sell their products or search other ads to buy.
During the 3 months of development, we implemented version control using GIT for seamless collaboration within the development team.
Successfully developed a robust e-Market web application hosted on Microsoft Azure.
Application for Android
Language Used: Java, SQLite
Created an Android application using Android Studio as an extension of the personal blog, enabling users to access piano notes for 250+ songs
Engineered the backend using Node.js to push to a JSON file hosted on Firebase, which is fetched in the app to automatically update its local SQLite database
Integrated Google AdMob to display ads to users requesting full piano notes, creating a monetization stream
Designed a responsive and minimal UI using RecyclerView and custom XML layouts to display song listings efficiently
