SAW SAW S S MART MART A A LLOCATION LLOCATION W W IZARD IZARD By: SHASHANK R. CHAUDHARI SHASHANK K.
SASAWWSSMART MART AALLOCATION LLOCATION WWIZARDIZARD
By:SHASHANK R. CHAUDHARISHASHANK K. DHAUNDIYAL
PROBLEM STATEMENT
Automatic Allocation Of Applicants To Exam Centers In A Calibrated Manner Considering Several Different Parameters
DELIVERABLE
To devise an algorithm for automatic allocation
To build a prototype to demonstrate the algorithm
INTRODUCTION TO SAW: SMART ALLOCATION WIZARD will be used during
Allocation Management where in the system using the applicant and the LISP database will be responsible to automate the allocation of LISP to the candidates according to their preferences based on various internal and external parameters.
It’s a web base application developed using C#, Asp.net & SQL
It’s a highly flexible system which can be used for all sorts of assessments and the records of these assessments could be stored in the database made available for future references
OVERVIEW: The approach used in developing SAW is based
on Rapid Application Development(RAD) Model reason being
i. Emphasis is on an extremely short development cycle
ii. Input and Output for the system are well understood
iii. Requirements are clear and the project scope is constrained
iv. Focus is on following a linear sequential form of development in an iterative manner
THE RAD PROCESS MODEL IS ILLUSTRATED IN THE FIGURE BELOW:
PROCESS INVOLVEDProcess for developing SAW is well defined after the input and the output requirements are cleared.
INPUT REQUIRED: Applicant list:
The applicant list is the list of registered applicants for an examination which has all the details in a format required by the system
Lisp list:Based on the assessment details and the rules involved. this would be a detailed list
Assessment details (from the customer):The details include date for the examination, no. of batches & timing of these batches for an assessment
Rules for the assessment (from the customer)These include rules to be followed for an assessment might be w.r.t allocation, scheduling, etc
PROCESS:
PROCESS 1: APPLICANT SORTING The applicants registered for an Assessment fill in their
detailed information as per the requirement for an assessment for example; app_name, address, gender, physical disability (if applicable), City, State, Country, Centre preference etc.
The sorting is done in 4 major parts:i.)Applicant list is segregated based on the cities of the
applicant using which the number of LISPs to be used for that assessment in that city could be decided.The output of this would be a table containing the city with its count of registered applicants
CITY NO. OF APPLICANTS
MUMBAI 5000
DELHI 8000
CONTD..ii.) The segregated APP list city-wise is then further sorted into: Physically handicapped, females and males.
The SAW system facilitates the scalability of maintaining various PH categories in the database. The same can be modified or deleted as per requirement.
This categorization will help in allocating a relevant LISP to a PH applicant, considering the applicant’s PH class.
The system then segregates the list further into PH female applicants and then PH Male applicant.
This process involves two database tables: LOOKUP_PH_TBL
MASTER_APPLICANT_TBL
CONTD..iii.) After sorting the PH app the system then clubs all the non PH Female applicants followed by the male applicants.
iv.) The processed applicant list after running through all the above mentioned parameters is displayed and it is this list which is used for the further procedure of applicant to LISP mapping/allocation.
PROCESS 2: LISP SORTING The process of LISP sorting is done with reference
to their availability for the assessments in terms of whether the LISP is available for those dates, timings & batches of an assessment.
The purpose of this process is to list down the available LISPs from higher-level to lower-level of order of their profitability
In order to accomplish this task, 5 parameters are to be considered
1.CAPACITY: The intake capacity of the LISP in accordance to its
workstation availability. (The higher, the better)
2.COST: Cost-per-workstation stated by the LISP. There are various
other cost heads that can be classified into pre-assessment and post-assessment cost heads.
3.ASSESSMENT EXPERIENCE: Recent experience an LISP has; in conducting assessments.
The input for this parameter is in the form of no. of applicants assessed
4.INFRASTRUCTURE: This parameter tracks all the facilities owned by an LISP, that
facilitate ease of operation
5.SUPPORT: Determines the level of support and co-operation received
from the top-level management & the employees of an LISP, in conducting an assessment
CALCULATION OF COST PER APPLICANT FOR LISP This is a sub process of LISP sorting in which the
cost incurred by an lisp per assessment is calculated to get the cost per applicant
This section tracks a wide spectrum of costs incurred before, during and after conducting an assessment. An LISP is cost-effective, if its cost-per-applicant is low.
This section stores expense-specific details of all the LISPs across multiple assessments, helping the user in calculating the cost-per-applicant and rating the LISP in terms of cost.
COST PROCESS FLOW
Update LISP parameter data: Each LISP has parameter fields
These fields should contain appropriate data (accurate and up-to-date). This data will enable the system to calculate a rating for each parameter by mapping the parameter values to the corresponding look-up tables.
The values for one or more LISP parameter may change across different assessments.
The SAW system is flexible enough to accommodate the new values while preserving the old ones for the purpose of future data analysis and decision making.
Table Structure In this process, DETAILS_LISPPARAMETERS_TBL is
used to store data. The Table Structure for the same is given below:
PROCESS 3: LISP RATING As the data of an LISP for its parameters are filled
in, it automatically gets updated for that LISP in the data base which helps us in tracking the benefits of using a particular LISP for future assessments.
A fixed weightage is predefined for an assessment for e.g.Node cap 0.15
Cost per node 0.25
Infra 0.1
Support 0.3
Experience 0.35
This process involves 5 database tables which are:These tables are basically lookup tables used in
giving rating to the LISP based on its available parameters
CALCULATING & STORING LISP SCORE Rating an LISP based on its Score is crucial to
determine its profitability and desirability.
The LISPs are rated 1-5 with 1 being the lowest score and 5 being the highest.
Rating process includes two sub-parts:a. Weightage in %b. Parameter Values
Parameter values are multiplied by the corresponding weightage values stored in the MASTER_ASSESSMENT table. The products are then added together to get the final LISP Score.
Table Structure The table which stores LISP Score-specific
information is DETAILS_LISPSCORE_TBL. The structure for the same is given below:
The ratings given to an LISP is then combined with the weightages to those parameters w.r.t the assessment conducted which gives us the weighted average score of an LISP
The LISPS are then sorted in their descending order of score which would be used for the allocation process
Purpose: These ratings help us in using an LISP in the best possible manner being more profitable to the organisation in terms of Cost, Operational ease, and better applicant experience
PROCESS 4: GOOGLE API MATRIX This process calculates the distance between an
applicant current location and all the selected LISPs to find an LISP closest to the applicant’s location.
It compares the applicant address fields of APP_AREA, APP_TOWN, APP_CITY, APP_PINCODE, APP_STATE and APP_COUNTRY to the corresponding address fields of an LISP
This process connects to the Google API Distance
Matrix service on the Google Server to find the distance (in KM) between the two addresses.
The calculated one-to-many distances against an applicant are stored in the database for future use and reference, to reduce the no. of pings on the google server.
Three tables are involved in this process, they are: MASTER_APPLICANT_TBL MASTER_LISP_TBL MASTER_DISTANCE_TBL
The table structure for MASTER_DISTANCE_TBL is as follows:
PROCESS 5: ALLOCATION PROCESS This process involves assigning a closest LISP which has
seats available for assessment to an applicant. The date and batch information is stored in the MASTER_APPLICANT_TBL
The applicants are 1st allotted the 1st shift of an lisp and then the 2nd shift is filled.
After both the shifts of the lisp are filled the system then starts mapping the applicants to the next closest lisp
With this there is reduction by 1 in the applicant list which in turn reaches 0 indicating all applicants have being allocaated successfully.
OUTPUT OF THE PROCESS Applicant list updated with the lisp id which is
allotted to the app
Also the details of the batch and timings of the app for that assessment at that respective lisp
CONCLUSION
RECOMMENDATION