Academy & Industry Research Collaboration Center (AIRCC)

Volume 12, Number 07, April 2022

A Data-Driven Real-Time Analytical Framework with Improved Granularity
using Machine Learning and Big Data Analysis


Yubo Zhang1 and Yu Sun2, 1Shenzhen College of International Education, China, 2California State Polytechnic University, USA


During daily studying and working, people have to research massive amounts of information on the internet and download numerous files. Some of these files can be easily categorized to relevant files. However, there are always some files left unorganized due to their difficulty in categorization [1]. Such files pile up in the download folder as time passes, making the folder extremely messy. Many people do not have the motive to clean up the folder as it requires a lot of energy and time. Based on this common problem, my group developed an app that can clean the messy folder up. After applying our program which is based on machine learning, the files will firstly be divided into five general parts – document, video, music, photos and package. Then the files will be further categorized based on the contents they present. For example, photos are divided into animals, families and so on. In order to achieve the content-categorizing function, several powerful apis were introduced in our program, and they helped us to achieve the optimal results.


Data Processing, Deep Learning, Machine learning.