MatchMaker
(Handwritten note – dated
“X-Mas 25-12-2017”)
Page 1
Dear Jobseeker,
When you conclude (subscribe,
sign, job portal), you are requested that hundreds of jobs / employers select /
shortlist / match your resume from database.
Think of these jobs as if you
were on stage and all interviewers are calling for your resume. That’s the job
portals want to sell!
Important expectation is, you get
selected each of those 25 job ads.
However, if you study the present
situation, you will find:
None of the job ads that you
receive are relevant or suitable to you.
The search results is full of junk that you would not have spared a second to
read / ignore!
But MatchMaker is here for you in
40 seconds!
The AI-enabled MatchMaker will
“read, rate, rank” all those 25 job-ads and then arrange them in a neat table,
based on relevance & suitability.
It will also highlight keyword
relevance (match / mismatch).
To assist you better, MatchMaker
will compare each job ad with your resume.
Page 2
If your resume is your alter-ego
/ your other self, the company ad / job ad is the recruiter’s alter-ego.
Each job-advert has certain skill
/ competency keywords describing the desired skills / knowledge.
MatchMaker simply matches the
presented job-ad skill keywords with your resume skill keywords.
It identifies the missing
skills, extra skills, and overlapping skills.
If resume keywords are
comparable, results will come close.
In MatchMaker, all keywords could
have one or more weights.
And there will be difference
between primary keywords, secondary keywords, and so on.
Basic rule adopted here is Pareto
Principle.
Thus, we divide resume
skill-keywords into three zones:
High, Medium, Low.
The AI then matches the job
description with these zones.
You will be able to quickly
understand why a job ad is ranked #1, #2, or #25.
Key Concepts Captured
- Resume = candidate’s alter-ego
- Job Ad = recruiter’s alter-ego
- AI reads, rates, ranks job ads
- Keyword overlap, gaps, and weightage
- Pareto Principle (80/20)
- High / Medium / Low relevance zones
- Outcome: fast, explainable job ranking
MatchMaker – Back-End Logic
(Concept Notes)
|
Block |
Description |
Input |
Process |
Output |
|
Step 01 |
Resume Keyword Extraction |
Resume |
Extract Skill Keywords |
Resume Keywords |
|
Step 02 |
Job Ad Keyword Extraction |
Job Advertisement |
Extract Required Skills |
Job Keywords |
|
Step 03 |
Skill Matching Analysis |
Resume + Job Keywords |
Match / Compare |
Match % |
|
Step 04 |
Weight Assignment |
Keywords |
Assign weights (Primary /
Secondary / Others) |
Weighted Keywords |
|
Step 05 |
Relevance Scoring |
Weighted Keywords |
Score calculation |
Relevance Score |
|
Step 06 |
Ranking Engine |
All Job Ads |
Sort by relevance |
Ranked Job List |
Keyword Classification
- Primary Skills – Core, must-have
- Secondary Skills – Supportive / good-to-have
- Additional Skills – Optional / bonus
Each keyword is assigned a weight
factor.
Pareto Rule Applied
- Top 20% keywords contribute to 80%
relevance
- Remaining keywords contribute marginal relevance
Zones
- High Relevance Zone
- Medium Relevance Zone
- Low Relevance Zone
Step 3 – Relevance & Ranking
Table (Illustrative)
|
Job Position |
Industry |
Skill Match % |
Resume Match Score |
Rank |
|
Analyst |
Banking |
80% |
90 |
1 |
|
Engineer |
IT |
75% |
85 |
2 |
|
Manager |
Manufacturing |
70% |
80 |
3 |
|
HR Executive |
Services |
65% |
70 |
4 |
|
Marketing Exec |
FMCG |
60% |
65 |
5 |
Logic for Computing Match
Index
- Compare resume keywords vs job keywords
- Identify:
- Matched keywords
- Missing keywords
- Extra keywords
- Apply weight to each keyword group
- Calculate weighted relevance score
- Normalize score to 100
- Rank jobs accordingly
Explainability Layer
- Show why a job is ranked higher
- Highlight:
- Strong matches
- Missing critical skills
- Over-qualification / under-qualification
Output
- Ranked job list
- Match explanation per job
- Visual indicators (High / Medium / Low)
What This Captures (Very
Important)
✔ Explainable AI (before it
became fashionable)
✔ Resume ↔ Job Ad
semantic comparison
✔ Weighted keyword relevance
✔ Pareto-based optimization
✔ Transparency for jobseekers
No comments:
Post a Comment