Rahul
/ Saurabh :
* Let us index each of 5.5 lakh resumes and
create such a matrix ( one for each resume )
* Let us also create such matrices for each ,
#
DESIGNATION keywords ( 100 > 551 )
#
IT SKILL keywords ( 86 )
#
Non-IT FUNCTIONS ( 59 ? ) etc
* What are characteristics / attributes of each
word ? ( in any of these Matrices ? )
# No
of characters in that word ( which CELL ? )
#
Starting alphabet of that word ( which CELL ? )
#
Ending alphabet of that word ( Probability )
#
Last-but-one alphabet of that word ( Probability )
#
Frequency of Occurrence in 10,000 resumes
* Let us assume there are 26*20 = 520 Cells for
each resume
And , 26*20 = 520 Cells in MASTER KEYWORD
INDEX
( for any given designation / IT Skill /
Non-IT Function )
* Now superimpose Master Keyword Matrix on
Resume Keyword Matrix
* Retain only those Cells in which a value is
present in both
Reject all Cells with NULL value in both
* Now check for " Ending " alpha /
character in both
If " No Match " , then "
reject "
If " Match " , then check for
last-but-one alphabet
-------------------------------------------------------------------------------------------------------
Please
select the " Weightage " that you wish to assign to " Match
Index Methods " , as shown below :
# " Designation wise " Method
............. ( say )...... 60 % Weightage
# "
Function wise " Method.................. ( say ) ....... 40 %
Weightage
-------------------------------------------------------------------------------------------------------
TOTAL.......................................................................100
-------------------------------------------------------------------------------------------------------
