Transcript Slide 1

UC Berkeley Undergraduate
Orientation Database
Project Review 3 – December 5, 2008
Group 5:
Peter Chang • Eric Follis • Justin Hsu • Jason Tan • James Tong
IEOR 115: Industrial and Commercial Data Systems
University of California, Berkeley
Mission Statement
"New students' initial encounters with the institution may have profound
effects on subsequent levels of involvement, and these encounters
should be carefully designed to socialize students to the institution's
highest educational values and goals."
(How College Affects Students, 1991)
UC Berkeley Undergraduate Orientation Database
CalSO, New Student Orientation
Design a comprehensive, all-encompassing
database to facilitate the operation of CalSO
Student Information
Counselor Information
Event Information
UC Berkeley Undergraduate Orientation Database
UC Berkeley New Student Services
(0,N)
Is_Accessible_For
Special Need
(0,N)
(0,N)
Requires
(0,N)
(0,N)
Campus
Location
(0,N)
Located_at
Financial Aid
(0,N)
Accommodates_for
(1,N)
Reserves
Is_Prerequisite_for
Requires
(0,N)
(0,N)
Is_Assigned_To
(0,N)
CalSO Event
(1,N)
Training
(0,N)
(1,N)
Creates
Lives_in
(0,N)
Attends
(0,N)
(0,N)
Dormitory
(0,1)
(0,N)
Lives_in
(0,N)
Lives_in
(0,N)
Major
(0,N)
(0,1)
Enrolls_in
(0,N)
(0,1)
Liaison_to
Supervises
Enrolls_in
(0,N)
(0,N)
(0,N)
(0,N)
(1,N)
Receives
(0,1)
(0,N)
(0,N)
(0,N)
(0,N)
(0,N)
(0,N)
(0,N)
Conducts
Coordinator
(0,N)
(0,N)
Counselor
(1,N)
(0,N)
(1,N)
New Student
Declares
(1,1)
Supervises
(1,N)
Tour
Director
Attends
(0,N)
O,P
Parent
Advisor
(0,N)
Speaker
(0,N)
Advises
Speaks_at
People
UC Berkeley Undergraduate Orientation Database
Relational Schema
UC Berkeley Undergraduate Orientation Database
Relationship View – MS Access
UC Berkeley Undergraduate Orientation Database
Add New People
UC Berkeley Undergraduate Orientation Database
Add New Event
UC Berkeley Undergraduate Orientation Database
Report: Performance Ratings
UC Berkeley Undergraduate Orientation Database
Normalization Analysis: 1NF
R is in 1NF if all attribute domains include only values that
are atomic (indivisible) and single-valued.
1NF:
Training(Training_ID, Name, Hours, Prerequisites, Required)
TrainingName(Training_ID, Name, Hours, Required)
TrainingPrerequisites(Training_ID, Prerequisite)
UC Berkeley Undergraduate Orientation Database
Normalization Analysis: 2NF
R is in 2NF if it is in 1NF, and every non-prime attribute is
fully functionally dependent on the Primary Key
2NF:
TourBuilding(Counselor_ID, CFname, CLname, Tour_ID,
TourName, Location_ID, Building)
CounselorName (Counselor_ID, CFname, CLname)
TourID (Tour_ID, TourName)
Location (Location_ID, Building)
TourLocation (Counselor_ID, Tour_ID, Location_ID, Building)
UC Berkeley Undergraduate Orientation Database
Normalization Analysis: 3NF
R is in 3NF if R is in 2NF and non-prime attributes of R are
transitively dependent on the primary key
3NF:
TourLoc (Counselor_ID, Tour_ID, Location_ID, Building)
LocationID(Location_ID, Building)
TourLocation (Counselor_ID, Tour_ID, Location_ID)
UC Berkeley Undergraduate Orientation Database
Query 1: Absence Analysis
In order to improve student outreach effectiveness, find the economic &
geographical demographic information of students who did not attend
CalSO
Assumptions:
•NewStudent.Attended = 1 if student attended CalSO, 0 if not.
•Economic & geographical demographic information is fully described by county of
residence and financial aid status.
SELECT
P.FName, P.MName, P.LName, P.Email, P.Phone, P.Major,
NS.Financial_Aid, P.Class_Standing
FROM
NewStudent as NS, People as P
WHERE
P.PID = NS.PID, NS.Attended = 0
GROUP BY P.County, NS.Financial_Aid;
UC Berkeley Undergraduate Orientation Database
Report: Non Attendees
UC Berkeley Undergraduate Orientation Database
Query 2: Event Effectiveness
Find the interest level associated with each optional event in CalSO.
Assumptions:
•Students are required to attend at least one event of their choice.
•Interest level is determined based on a weighted function of the CalSO event
attendance level, event survey score, and the counselor performance score.
•InterestLevel = 4*AttendanceRate + 6*P.Score + 3*S.Score.
SELECT
CE.Event_ID, PP.Year, SUM(CE.No_of_Attendee) / COUNT(S.NewStudent)
as AttendanceRate, P.Score, 4*AttendanceRate + 6*P.Score +
3*S.Score as InterestLevel
FROM
CalSO_Event as CE, Survey_Rating as S, Counselor as C, NewStudent
as NS, Performance_Rating as P, People as PP
WHERE
CE.Event_ID = S.Event_ID, C.Counselor_ID = CE.Counselor_ID,
C.Counselor_ID = NS.Counselor_ID, P.Counselor_ID = C.Counselor_ID
GROUP BY CE.Event_ID, PP.Year;
UC Berkeley Undergraduate Orientation Database
Query 3: Parents Attendance Forecast
Forecast parent attendance for CalSO events using regression analysis in
order to optimize the resource allocation for future years.
Assumptions:
•There exists a relationship between the attendance levels of students and the
attendance level of parents.
•Regression Formula: y=Xβ+ε
•X = # students in attendance
•Y = # students in attendance
•SQL code below extracts necessary inputs to perform the proposed regression
analysis. All calculations will be executed in MS Excel using macros written in Visual
Basic for Applications (VBA).
SELECT
FROM
WHERE
GROUP BY
COUNT(Pa.PID), COUNT(NS.PID), NS.Sem_Admit, PP.Year
People as PP, Parent as Pa, NewStudent as NS
PP.PID = Pa.PID, Pa.Student_SID = NS.Student_SID
PP.Year;
UC Berkeley Undergraduate Orientation Database
Query 4: Correlation of Training & Performance Review
What is the correlation between the amount of optional training received
and counselor performance?
Assumptions:
•Correlation Equation:
•X = # optional training hours received by counselor
•Y = performance score received by counselor
•SQL code below extracts necessary inputs to perform the proposed correlation
analysis. All calculations will be executed in MS Excel using macros written in Visual
Basic for Applications (VBA).
SELECT
FROM
WHERE
C.Counselor_ID, T.Training_ID, SUM(T.Hours), P.Score, P.Year
Peformance as PF, Training as T, Counselor as C, People P
P.PID = S.PID, PF.Counselor_ID = C.Counselor_ID, T.Training_ID =
S.Training_ID, T.Required = ‘No’
GROUP BY C.Counselor_ID, P.Year;
UC Berkeley Undergraduate Orientation Database
Query 5: Optimal Number of Employees to Hire
Uses linear programming to determine optimal number of employees to
recruit and hire in order to meet student demand.
Assumptions:
•Optimality is defined as minimizing costs while meeting a required service levels.
•Coordinators do not contribute to the required student to staff employment ratio.
•SQL code below extracts necessary inputs to perform the proposed linear
programming analysis. All calculations will be executed in MS Excel using macros
written in Visual Basic for Applications (VBA).
UC Berkeley Undergraduate Orientation Database
Query 5: Optimal Number of Employees to Hire
The following linear program is used:
Decision Variables
X1 = # new hires
X2 = # experience hires
X3 = # coordinators
Fixed Variables
C1 = wage of new hires
C2 = wage of experienced hires
C3 = wage of coordinators
S = # new students
User Inputs
B = total budget
G = required counselor to coordinator ratio
R = required student to staff employment ratio
P = required percentage of experienced hires per total
number of hires
SELECT
FROM
WHERE
C.Wage, CO.Wage, CO.Experience, P.Year
People as P, Counselor as C, Coordinator as CO
P.PID = C.PID, C.Supervisor = CO.PID, C.Counselor_ID =
NS.Counselor_ID
GROUP BY P.Year, CO.Experience
UNION
SELECT
COUNT(NS.PID)
FROM
NewStudent as NS;
UC Berkeley Undergraduate Orientation Database
Query 5: Optimal Number of Employees to Hire
Access Output
Excel
UC Berkeley Undergraduate Orientation Database
Thank You
UC Berkeley Undergraduate Orientation Database