Michael Egan 

BSc [Hons] Creative Computing

I am a 29 year old mature student from Dublin. In the last four years studying at IADT I have gained the knowledge and skills of a full stack developer and UI designer using multiple languages and frameworks. By undertaking professional practice I learned how to work in a team to meet fortnightly deadlines. I also particularly enjoyed visually displaying and interpreting data using programs such as Tableau and Data Desk.

The Beaten Trail

The application created for this project, is a travel-based, trip research and planning application – called the beaten trail. This application focuses on allowing users to plan a multi destination trip, by creating a trip with more than one destination. The user can personalise each trip created by adding text, hashtags and photos to a trip. Users can also use the beaten trail to research trips created by other users to discover popular routes and potentially gain inspiration for a future holiday. Finally, A social aspect is added to the application by allowing users to like other users’ trips, creating a sense of community in the beaten trail.


A Travel Web Application With a Recommender System

The focus of the literature review for my 4th year research project is on machine learning’s recommender systems. Recommender systems are a subfield of machine learning, which are designed to predict outcomes which will give the user the best content that is personal to them by using calculations based on user feedback.
A memory-based collaborative filtering recommender system is implemented into the beaten trail to predict and recommend destinations a user may like to visit when creating a trip.

Memory-based was chosen as it is the more popular method to implement rather than model-based.Collaborative filtering was chosen to allow a user to discover new interests based on similar user preferences, despite the the potential of a cold start issue with collaborative filtering. The recommender system was created as a Flask web application which runs the Python script to calculate the recommendations and acts as an external API when called by the React frontend of the main MERN stack application to retrieve and display the recommendation to the user of the beaten trail.