Computer Aided Academic Planning

1
Computer Aided Academic Planning Scheduling

• X. Fu, B. Peltsverger, C. L. Rozmus
• Georgia Southwestern State University
• 2005 SACS-COC Annual Meeting

2
What is CACPS

• Computer Aided Curriculum Planning Scheduling
  System
• Pilot project at Georgia Southwestern State
• Software System
• Helps students make academic plan
• Helps advisors monitor academic progress
• Helps administrators optimize resource util.

3
Personalized Study Plan

• CACPS relies on Personalized Study Plan
• What classes to take?
• When to take them?
• Not a sample 4-year study plan!
• Caters to students personal interests
• Adaptable to Multi-year Course Offering

4
What Does CACPS Do?

• Automatically generates study plans
• Facilitate advisor/advisee interaction
• Information collection
• Helps smart decision
• Like smart order system for retailers

5
Outline

• Why automatic academic planning?
• How to plan?
• Tool Demo
• Dissemination Future Directions
• Discussion (10 min)

6
Motivation
7
Challenges Higher Education

• Diversified student body
• More part-time students
• Increasing student average-age
• ? Diversified academic backgrounds, goals
• Limitation of educational resources
• Budget cuts
• Flow of faculty members
• ? Frequent changes of course offering

8
Concern of Students

• How to control academic progress?
• Given the career goal
• What classes to take?
• When to take them?
• What if course offering plan is changed?

9
Concern of Advisors

• Improve advising service efficiency
• How to retrieve information effectively?
• Given student transcripts, how to find out
  problems quickly?

10
Concern of Administrators

• Better service with better resource utilization?
• Which classes are most desired?
• Which classes are outdated?
• Which classes must be offered not to jeopardize
  graduation plan of students?

11
Solution Automatic Planning Tool

• Students
• Advisors
• Administrators

12
Prototype Application at GSW

• Input
• Student Preference
• Transcripts
• Degree Requirements
• Course Offering Schedule
• Course Description
• Pre-requisite Relationship
• Output
• Customizable Study Plan

13
System Architecture
GSW Banner
Degree Audit System
Text Parser/Processor
Planner
Data Store
14
User Manual

• Step 1 Specify Preference
• Max hrs per semester, etc.
• Step 2 Curriculum Planning
• What classes to take?
• Step 3 Plan Verification
• All degree requirements met?
• Step 4 Plan Generation
• Plan is automatically generated
• Step 5 Plan Refinement
• Use interaction interface

15
Tool Demo
16
Demo Schedule

• Start App. (0.5 min)
• Show Input Files (1 min)
• Input Preference (1 min)
• Course Selection (2 min)
• Modify preference (1 min)
• Generate Study Plan (1 min)
• Adjust Study Plan (1 min)
• Another Senior Student (1.5 min)
• Total 10 min

17
Future Directions
18
Current Status

• Two technical challenges already solved
• How to collect information?
• Mainly parse from existing web pages
• Pros very low development cost
• Cons not adaptable to changes of web pages
• Directions more DB adaptor
• How to automatically generate academic plan?
• A modified topological sort
• Directions introduction of math programming

19
Extension

• Broader Impacts
• Windows app. vs. Web app.
• Retain ease of use
• Push vs. Pull
• More adaptable to changes
• Dissemination
• GPL vs. Proprietary Software
• Customize to different institutions
• More general template of degree requirements,
  transcripts
• Authoring tools
• More DB connector modules

20
A Bigger Picture
DONE!
Study Planner
DB
21
Optimize Course Offering

• Big problem small liberal art institutions
• How to optimize educational resources?
• Decision making very tricky!
• Example
• If Spring 2006 offering of Computer Architecture
  is cancelled ? Any graduation plan affected?
• Simply query the Study Plan DB!

22
More Ambitious Plan

• Automatic optimizer of multi-year course offering
• Input
• Study plans of all students
• Capacity of each department
• Current Course Offering Plan
• Output
• Improved Course Offering Plan
• Balanced class size
• Improved Student Satisfaction Rate
• Graduation date
• of class cancellation

23
Design Idea

• Start from current offering plan
• Random adjustment ? A new plan
• Evaluate new offering plan
• If better, enlist the new plan
• Repeat above procedure until an optimal plan is
  found or pre-set time limit is met

24
Optimize Scheduling ofClass Time

• Optimized multi-year course offering plan does
  not solve ALL the problem!

I can NOT take both of my Computer Programming
and English Composition classes! Time Conflicts!
25
Class-Time Scheduling System

• Input
• Study plans of all students in a semester
• Output
• Class scheduling (i.e., which class is offered at
  which time)
• Objective
• Minimize the of time conflicts

26
Design Idea

• Modeled as quadratic programming prob.
• Class-time Schedule Allocation Matrix
• Schedule Cost Cost Matrix
• Linear on Allocation Matrix
• Total Cost product of two matrices
• Objective valuation of Allocation Matrix s.t.
  total cost is minimal

27
Conclusion

• Challenge of diversity
• Personalized Study plan
• Monitor academic progress
• Demo tool
• Automatic plan generation
• Interaction tool for advisor
• Future plans
• Optimize class offering
• Optimize class scheduling

28
Opportunities

• We are looking for opportunities
• Funding
• Collaboration
• GSW Web-i-Tech

29
Questions?