Improved Recommendation System Based On Play Time
Disciplines
Computer Engineering | Digital Communications and Networking | Other Computer Engineering
Abstract (300 words maximum)
The recommendation system has been utilized for many years with many companies and studied across the machine learning field. Many have used this algorithm to try and give the user suggestions based on what they have watched, played, and read across the internet. From YouTube, X (Formally Twitter), Google, and many other companies, this system has been used to the very limit. However, there is always one problem when it comes to this sort of system. It is the way that it has been implemented and the way it understands what the user is into at the current moment and tries to recommend something that ultimately does not fall in line with what the user’s current interests are at that moment.
This project that is still in the works is to help improve the accuracy of these recommendations so that users using this system can be recommended things that align with their current interests. The platform this will be based on will be with the Steam PC Market platform. Steam is one of the largest PC Marketplaces in the world. From games to even business software like Blender and PowerDirector, it has all kinds of software and even hardware that users will be interested in looking at and potentially buying as well. This project will be using a Neural Network to train based off a user’s time with an application with the use of Steam’s API, because Steam actively tracks a user’s time inside an application and keeps a log of that in their system. This will help improve the recommendation system greatly as it will continuously train on the time spent on an application but also learn what the user likes and does not like.
Academic department under which the project should be listed
CCSE - Computer Science
Primary Investigator (PI) Name
Dr. Md Abdullah Al Hafiz Khan
Improved Recommendation System Based On Play Time
The recommendation system has been utilized for many years with many companies and studied across the machine learning field. Many have used this algorithm to try and give the user suggestions based on what they have watched, played, and read across the internet. From YouTube, X (Formally Twitter), Google, and many other companies, this system has been used to the very limit. However, there is always one problem when it comes to this sort of system. It is the way that it has been implemented and the way it understands what the user is into at the current moment and tries to recommend something that ultimately does not fall in line with what the user’s current interests are at that moment.
This project that is still in the works is to help improve the accuracy of these recommendations so that users using this system can be recommended things that align with their current interests. The platform this will be based on will be with the Steam PC Market platform. Steam is one of the largest PC Marketplaces in the world. From games to even business software like Blender and PowerDirector, it has all kinds of software and even hardware that users will be interested in looking at and potentially buying as well. This project will be using a Neural Network to train based off a user’s time with an application with the use of Steam’s API, because Steam actively tracks a user’s time inside an application and keeps a log of that in their system. This will help improve the recommendation system greatly as it will continuously train on the time spent on an application but also learn what the user likes and does not like.