Globalcert Data Science Internship TASK-2: Week-2 Analysis

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 14

GLOBALCERT DATA SCIENCE

INTERNSHIP
TASK-2
Week-2 Analysis
AUTOMATED SYSTEM FOR APPLICANT
SELECTION
Presented by
TEAM-4
PROBLEM DEFINITION
GlobalCert is looking for full-time employees for a variety of roles with different pay levels. They
want to extend their firm into different dynamics as a fledgling start-up, which demands recruiting
extremely competent and acceptable employees to run their firm. Furthermore, they need your help
as they lack a sufficient automated system to optimize their hiring process.
ABOUT THE DATASET
There is quite a number of columns we have considered in our dataset.
By using the method of open source data we have constructed our own database with the following
columns-

Name Bachelors degree and percentage

Contact Number History of Backlogs

Gender English test percentage

Age Department applying to

10th Board and percentage If MBA, percentage

12th Board and percentage Work experience

12th field Salary expectations per month


APPROACH
COMPANY AUTOMATED SYSTEM
Filling in the google form or applying on the Data gathering using the open source data
website method

Details get stored on the database Splitting the data department-wise

Processing the data using the automated system Processing the data department wise using
build Quartile Method

The potential applicants move on to the later


Displaying the top 10 potential applicants
interview rounds
ANALYSIS
Analysis is performed using the following parameters-
1. 10th,12th and degree percentage
2. MBA grade
3. History of backlogs
4. Work experience

Departments considered-

Customer
Marketing
Teaching Service
and HR

Marketing Technical
and Finance
GENERAL ANALYSIS

• Maximum no of applicants have 0 and 1


year of work-experience

• Gender ratio of the applicants


• Around 1600 male and 1400 female applicants
( depends on the situation to maintain the gender ratio)
• Maximum applicants have pursued Btech while the least with MCOM and MED

• Maximum applicants have studied their SSC in Central • Maximum people have studied their HSC from other
boards
1. MARKETING AND HR
Considering the factors ( 10th grade, 12th grade, degree
grade and History of backlogs) we have the final graph as
given below-

0-20k=1
20k-40k=2
40k-60k=3
60k-1L=4
1L-1.2L=5

Displaying top potential applicants with Final_Part: 343,344,443,444 (quartile method- a


part of RFM method)
2. MARKETING AND FINANCE
Considering the factors ( 10th grade, 12th grade, degree
grade, and History of backlogs) we have the final graph as
given below-

Displaying top potential applicants with Final_Part: 344,444,343,443 (quartile method)


3. TEACHING
Considering the factors ( 10th grade, 12th grade, degree
grade and History of backlogs we have the final graph as
given below-

Displaying top potential applicants with Final_Part: 344, 444 (quartile method)
4. TECHNICAL
Considering the factors ( 10th grade, 12th grade, degree
grade and History of backlogs) we have the final graph as
given below-

Displaying top potential applicants with Final_Part: 343, 344, 443, 444 (quartile method)
5. CUSTOMER SERVICE
Considering the factors ( 10th grade, 12th grade, degree
grade and History of backlogs) we have the final graph as
given below-

Displaying top potential applicants with Final_Part: 132, 133, 134, 142, 143, 144, 232, 233,
234, 242, 243, 244, 332, 333, 334, 342, 343, 344 (quartile method)
CONCLUSION
GloberCert received a load of applications in various departments
The automated system is successfully build in a way that, depending on the major factors(10 th, 12th, and degree
grades, MBA grades if required , history of backlogs and Work-experience the system automatically displays a list of
top applicants who should be selected.

10 applicants from each department with the potential to join the department are found. They can be called for the
round 2 of the interview for more screening and deciding upon the job profiles.
THANKYOU
- TEAM-4

You might also like