5.py

# -*- coding: utf-8 -*-
"""Job 34 Validate career recommendations.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1JTz0QznGN0CC6AaP1XFzLhIu6tL5p3vb

Validate career recommendations

Confirm Data Integrity
Before making recommendations:

Check for duplicates: Many entries repeat (e.g. repeated skills for multiple careers).
→ Remove duplicates by Career + Domain + Skill/Sub-skill.
Normalize skill names: e.g. “API Consumption”, “APIs API Design” should be standardized (“API Design”).
Validate thresholds and weights: Each “Threshold” (0–100) pairs with a difficulty level; ensure consistent mappings (e.g. Easy = 40.0, Medium = 60.0, Hard = 75.0).

Define Validation Criteria
You can validate a career recommendation (target career) using one or more of the following feature alignments with; Validation, Metric	Description	and Example.

Skill overlap (%) -  % of skills the user already has vs. expected for the target career. For example;	If the Data Engineer role requires 40 skills, user has 28 → 70% readiness

Average difficulty match -  Compare user’s level vs. Level Required. For example; If user’s average level < required → recommend learning path before transition

Domain fit -  Compare user’s current domains (Programming, Data, Security, etc.) to target. For example; Helps detect career/domain alignment

Tag similarity -	 Match interest/experience keywords in Tags. For example;	E.g. “Streaming, ETL” aligns with Data Engineer but not UX Designer

Threshold-weighted readiness -	 Weighted scoring using the Weight and Threshold columns for skill importance. For example; Weighted composite = ∑ (skill readiness × weight)

Example Validation Workflow
Let’s say you want to validate if someone should transition to Data Engineer:

Input (User Profile):

Current skills: SQL, Python scripting, CI/CD Basics, AuthN/AuthZ, Unit Testing.
Current Role: Backend Developer
Validation Steps:

1. Filter dataset → Career = 'Data Engineer'
2. Extract target skills and their Level Required.
3. Compute overlap with user’s skills:
* SQL ✓
* Python scripting ✓
* CI/CD ✓
* AuthN/AuthZ ✓
* Unit Testing ✓
→ 5/5 match → readiness = 100%
4. Compute difficulty gap: user skill levels >= required → no gap.
5. Weighted readiness = (∑ weights of matched skills / total career weight) → if > 0.6 → indicate good readiness to transition.

Implement Validation
"""

import pandas as pd

# load dataset
df = pd.read_excel('New jobready_career_skill_dataset_1000_rows_v2.xlsx')

def validate_career(user_skills, target_career):
    career_df = df[df['Career'] == target_career]
    skills_required = set(career_df['Skill/Sub-skill'].str.lower())
    user_normalized = set(s.lower() for s in user_skills)

    matched = skills_required & user_normalized
    readiness_score = len(matched) / len(skills_required)
    threshold_mean = career_df['Threshold (0-100)'].mean()

    return {
        "career": target_career,
        "readiness": round(readiness_score * 100, 2),
        "avg_threshold": threshold_mean,
        "matched_skills": list(matched)
    }

# Example
validate_career(["SQL Basics", "Unit Testing", "CI/CD Basics"], "Data Engineer")

"""Evaluate with Domain Clusters
Use domain grouping (e.g., APIs & Integration, Programming, Quality & Testing) to enhance validation accuracy:

* High overlap in 3+ relevant domains → strong recommendation.
* Low overlap in core domains → suggest bridge learning path before recommending.
"""