Data collection
Gathers about 3,000 registration records, each with a student ID, course ID, current study level and enrollment time. Every record is checked for correctness and completeness.
InputValidatorLinkedList
CS210 · Prince Sultan University
Sorting 3,432 registrations, timed in ms.
An intelligent system for analysing students’ course registrations. Records are stored in a self-developed linked list, scored for course demand, and sorted with four algorithms that are timed against their theoretical complexity.
O(1) tail insert
Merge Sort · 17 ms
head → LinkedList<Registration>
size 0
format id;course;level;hour
3,432
Registrations read & validated
4
Sorting algorithms compared
17ms
Merge Sort on the full dataset
15
Java source files, one job each
00 · Abstract
Developing an intelligent system for students’ course registration analysis.
This project designed an intelligent system for managing students’ course registration records at Prince Sultan University. It predicts course demand using data structures such as linked lists and arrays, together with sorting algorithms.
Records are stored in a self-developed linked list, and four sorting algorithms (Selection Sort, Insertion Sort, Merge Sort and Quick Sort) are compared on their time-complexity performance.
The results matched theory. On a dataset of more than 3,000 records, Merge Sort and Quick Sort, both O(n log n), beat Selection Sort and Insertion Sort, both O(n²), by a wide margin. This confirms that choosing the right algorithm matters for managing large data efficiently.
01 · System design
Three phases turn raw registrations into demand insight.
The system takes student course registration records and works out how much demand each course has. The data is then sorted four times, and each sort is timed.
Processing
For each node, the demand scorer calculates a score from four criteria: a base score, a class multiplier, the time of registration, and the course ID priority.
DemandScorer
Documentation
Writes the results to a new file with each student’s demand score. The data is left unsorted so the sorting algorithms can be performance-tested afterwards.
FileService
02 · Watch them race
Same input, four algorithms, every comparison counted.
Each lane sorts the same shuffled array of 36 values in descending order, just like the project. They all advance at the same number of comparisons per frame, so the finishing order reflects how much work each algorithm does.
sorting-race · n = 36
Selection Sort O(n²)0 comparisons
Insertion Sort O(n²)0 comparisons
Merge Sort O(n log n)0 comparisons
Quick Sort O(n log n)0 comparisons
03 · Measured on 3,432 records
The O(n log n) sorts win.
On 3,432 registrations, Merge Sort finished in 17 ms and Quick Sort in 22 ms. Insertion Sort took 26 ms and Selection Sort took 45 ms, as their O(n²) complexity predicts.
sort time · milliseconds
O(n²)O(n log n)
04 · Explore the project
Everything from the design to the last line of code.
Introduction
Why sorting matters, and the question this project answers.
ReadSystem design
The three phases and an interactive demand-score calculator.
ExploreAsymptotic complexity
Line-by-line T(n) derivations for every class.
AnalyseInterpretation
Why the 3,432-record dataset matters to the results.
ReadTeamwork
Who built which classes and wrote which sections.
Meet the teamSource code
All 15 Java source files, highlighted and ready to copy.
BrowseRun screenshots
Console output from the full program run.
ViewConclusion
Findings, lessons learned and future work.
ReadCS210 · Section 951 · Dr. Najla Althuniyan
Read the full write-up, or run the code yourself.
The report covers the method, tables and discussion. The repository has the source and the input data.