MichaelYitzchak/LinkedInJobPostings

📊 LinkedIn Job Posting Engagement Analysis

Which LinkedIn job posting characteristics predict candidate engagement (views) — and how well can engagement be predicted or classified using only posting-level features?

Personal motivation: As someone in entrepreneurship, understanding which job posting features attract candidates is directly relevant to future hiring decisions.

---

📹 Presentation Video

<video src="https://huggingface.co/datasets/YOUR_USERNAME/YOUR_REPO/resolve/main/presentation.mp4" controls style="max-width:720px;">

---

🚀 Interactive Dashboard

👉 Open the LinkedIn Job Engagement Dashboard

Tab	Description
🎯 Engagement Predictor	Enter posting details → get predicted views + High/Normal classification in real time
📊 EDA Dashboard	All 5 EDA findings as interactive charts
ℹ️ About	Feature groups, model details, limitations

---

📋 Dataset at a Glance

Property	Value
Source	LinkedIn Job Postings — arshkon/linkedin-job-postings (Kaggle)
Original size	123,850 rows × 49 columns
Working sample	30,000 rows · random_state=42
After join with companies	30,000 rows × 40 columns
After cleaning	29,572 rows × 51 columns (in df_model)
Train / Test split	23,657 / 5,915 (80/20, random_state=42)
Regression target	log_views = log1p(views) — log-transformed to handle right skew
Classification target	high_engagement — top 25% of training views (threshold derived from training set only)

---

⚠️ Scope & Limitations

LinkedIn's algorithm, sponsored status, and company follower counts drive the majority of view variance and are unobservable in this dataset. Models use posting-level features only. The practical goal is ranking postings by predicted engagement, not exact point prediction. Results show associations, not causal relationships.

---

🗂️ Repository Files

File	Description
notebook.ipynb	Full pipeline: Cleaning → EDA → Feature Engineering → Clustering → Regression → Classification → Bonus
linkedin_regression_model.pkl	Winning regression model: Random Forest (Tuned via RandomizedSearchCV)
linkedin_classification_model.pkl	Winning classification model: Decision Tree (max_depth=8, class_weight="balanced")
regression_model_results.csv	Full regression model comparison table
classification_model_results.csv	Full classification model comparison table

---

🧹 Data Cleaning Pipeline

7 steps from 123,850 raw rows to a clean, leakage-free modelling matrix:

Step 1 — Reproducible sampling
        123,850 rows → sample(n=30,000, random_state=42)
        Joined with companies.csv on company_id (left join, rows preserved)
        Result: 30,000 rows × 40 columns

Step 2 — Duplicate & missing target removal
        Removed duplicate rows
        Dropped rows where views is NaN or negative
        Result: 29,572 usable rows

Step 3 — Date parsing
        listed_time, original_listed_time, expiry, closed_time → parsed to datetime
        Extracted: posting_year, posting_month, posting_dayofweek, posting_weekend

Step 4 — Missing value analysis & column dropping
        Threshold: >70% missing → drop
        Dropped: closed_time (99.2%), skills_desc (98.1%), med_salary (95.1%),
                 remote_allowed (87.9%), applies (81.1%), max_salary/min_salary (76%)

Step 5 — Leakage columns excluded
        expiry, applies → removed (post-publication outcomes)
        views → kept as target only, never as a feature

Step 6 — Salary imputation strategy
        has_salary_info = 1 if salary present, else 0
        salary_midpoint computed from min/max salary where available
        Missing salary → imputed inside sklearn Pipeline on training data only

Step 7 — Log transformation of target
        Raw views: mean=14.9, std=98.8, max=9,949 — heavily right-skewed
        log_views = log1p(views) — compresses scale, improves regression fit
        Predictions converted back via expm1() for interpretation
        Outliers (IQR method): 4,074 (13.8%) — kept, not removed

---

🔍 EDA — 5 Research Questions

Note on notebook ordering: Q1=Work type, Q2=Salary, Q3=Description, Q4=Day of week, Q5=Seniority. Presented below in order of business impact.

---

💰 Q2 — Salary Transparency vs Views

No salary info   ████████████░░░░░░░░░░░░░  ~12 avg views   (70.1% of postings)
Has salary info  ████████████████████████░  ~21 avg views   (29.9% of postings)
                                             +74.3% lift ✓

Only 8,562 of 29,572 postings (29.9%) disclose salary. Transparent postings attract 74.3% more views on average. This is the highest-leverage, lowest-cost recruiter action available.

---

📝 Q3 — Description Length vs Views

< 100 words    ██████░░░░░░░░░░░░░░  ~8 avg views   — signals incomplete posting
100–250 words  █████████░░░░░░░░░░░  ~13 avg views
250–500 words  ████████████████████  ~24 avg views  PEAK ★ — sweet spot
500–750 words  ████████████████░░░░  ~18 avg views
> 1000 words   ███████░░░░░░░░░░░░░  ~10 avg views  — overwhelms candidates

Non-linear relationship confirmed. Sweet spot: 250–500 words. This motivated description_density — the #1 feature in the winning regression model.

---

📅 Q4 — Day of Week vs Views

Monday    ████████████████████  39 avg views  ★ best day  (n=1,837)
Tuesday   █████████████████░░░  25 avg views
Wednesday ████████████████░░░░  22 avg views
Thursday  ███████████████░░░░░  18 avg views
Friday    ███░░░░░░░░░░░░░░░░░   7 avg views  ✗ worst day (n=10,076)
Saturday  ████████████░░░░░░░░  28 avg views  (weekend — n=2,116 total, noisier)
Sunday    ████████████░░░░░░░░  28 avg views  (weekend — noisier)

Counterintuitive finding: Weekend postings show higher averages (~28), but the weekend sample is tiny (2,116 obs total) making these estimates unreliable. Monday is the clear best weekday at 39 avg views. The day-of-week signal is modest and should not override content features.

---

💼 Q1 — Work Type vs Views

Contract    ████████████████████  29.97 avg views  median: 7.0
Internship  █████████████████░░░  25.71 avg views  median: 5.0
Full-time   ████████░░░░░░░░░░░░  13.70 avg views  median: 4.0  ← 80% of volume
Other       ███████░░░░░░░░░░░░░  11.27 avg views  median: 4.0
Part-time   ██████░░░░░░░░░░░░░░   9.59 avg views  median: 4.0

Contract and Internship roles show the highest engagement. However, Full-time dominates volume (23,674 of 29,572 postings = 80%). Work type is a useful predictive feature but should not be interpreted as causal.

---

🎓 Q5 — Seniority Level vs Views

Entry-level  ████████████████████  18 avg views  n=792
Senior-level ████████████░░░░░░░░  16 avg views  n=3,577
Other/Mid    ██████████░░░░░░░░░░  15 avg views  n=25,203

Entry vs Senior: +12.4% more views
Entry vs Other:  +18.9% more views

Supply-side effect — more candidates qualify for junior roles, so the pool is larger. is_entry_role carries predictive signal because it proxies for candidate pool size, not intrinsic desirability.

---

🔥 Feature Correlation with log(views+1)

Feature                      Corr    Direction   Note
─────────────────────────────────────────────────────────────────────
desc_salary_interaction      +0.18   ↑ views     strongest single predictor
has_salary_info              +0.14   ↑ views     salary transparency
salary_log                   +0.12   ↑ views     salary level
description_density          +0.10   ↑ views     content quality
description_word_count       +0.08   ↑ views     description length
is_software_role             +0.08   ↑ views     tech role demand
is_data_role                 +0.07   ↑ views     data role demand
is_entry_role                +0.06   ↑ views     larger candidate pool
posting_weekend              -0.04   ↓ views     small negative signal
is_senior_role               -0.03   ↓ views     smaller candidate pool
─────────────────────────────────────────────────────────────────────
Internal correlations (structural — not data leakage):
salary_log ↔ salary_midpoint  +0.96  log transform of same variable
desc_wc ↔ desc_density        +0.55  density uses length in formula
is_software ↔ is_data         +0.35  often co-occur in job titles
is_senior ↔ is_entry          -0.28  mutually exclusive by construction

Most features show weak linear correlation — no single feature dominates. This motivated tree-based models (Random Forest, Gradient Boosting) which capture non-linear interactions and feature combinations.

🌡️ Correlation Heatmap (feature-to-feature + target)

                          log   desc  has   sal   desc  is_   is_   is_   post  is_
                          views dens  sal   log   wc    soft  data  entr  wknd  snr
──────────────────────────────────────────────────────────────────────────────────────
log_views              │  1.00  0.10  0.14  0.12  0.08  0.08  0.07  0.06 -0.04 -0.03
description_density    │  0.10  1.00  0.02  0.04  0.55  0.01  0.01 -0.01  0.00  0.00
has_salary_info        │  0.14  0.02  1.00  0.72  0.03  0.06  0.07 -0.03 -0.01 -0.02
salary_log             │  0.12  0.04  0.72  1.00  0.04  0.05  0.06 -0.02 -0.01 -0.01
description_word_count │  0.08  0.55  0.03  0.04  1.00  0.01  0.01 -0.01  0.00  0.00
is_software_role       │  0.08  0.01  0.06  0.05  0.01  1.00  0.35 -0.08  0.00 -0.05
is_data_role           │  0.07  0.01  0.07  0.06  0.01  0.35  1.00 -0.06  0.00 -0.04
is_entry_role          │  0.06 -0.01 -0.03 -0.02 -0.01 -0.08 -0.06  1.00  0.01 -0.28
posting_weekend        │ -0.04  0.00 -0.01 -0.01  0.00  0.00  0.00  0.01  1.00  0.00
is_senior_role         │ -0.03  0.00 -0.02 -0.01  0.00 -0.05 -0.04 -0.28  0.00  1.00
──────────────────────────────────────────────────────────────────────────────────────
Key structural correlations:
  salary_log ↔ has_salary_info  +0.72  same underlying signal, different form
  desc_wc    ↔ desc_density     +0.55  density formula uses word count
  is_software ↔ is_data         +0.35  frequently co-occur in job titles
  is_entry   ↔ is_senior        -0.28  mutually exclusive flags

The heatmap confirms no multicollinearity crisis — the highest inter-feature correlation (salary_log ↔ has_salary_info at 0.72) is a structural relationship between two forms of the same signal, not a data problem. All correlations with log_views are weak, validating the move to non-linear tree-based models.

---

⚙️ Feature Engineering — 20 Base + 6 Cluster = 30 Total Features

Group	Features
Text length	title_length, title_word_count, description_length, description_word_count
Text structure	description_density ★, title_desc_ratio
Salary	salary_midpoint, salary_range, has_salary_info, salary_log
Role keywords	is_senior_role, is_entry_role, is_software_role, is_data_role, is_manager_role, is_sales_role, is_marketing_role, is_remote_text
Interactions	desc_salary_interaction ★, senior_salary, weekend_remote, title_desc_word_interaction, salary_density_interaction, salary_description_interaction, title_density_interaction
Clustering	cluster_0, cluster_1, cluster_2, cluster_3, cluster_4, cluster_5

Missing value strategy:

• Columns with >70% missing → dropped
• Salary → has_salary_info flag + salary_midpoint where available; remaining NaN imputed inside sklearn Pipeline on training data only
• Remaining numeric → SimpleImputer(strategy="median") inside Pipeline

---

🔵 Clustering — KMeans k=6

Features used for clustering (12 total, leakage-checked): title_word_count, description_word_count, salary_log, description_density, has_salary_info, is_senior_role, is_entry_role, is_software_role, is_data_role, is_manager_role, is_sales_role, is_marketing_role

Methods used to select k:

1. Elbow method — inconclusive, no sharp elbow
2. KMeans silhouette scores on full training matrix
3. Cluster-size stability table
4. Interactive K-Means widget (visualization aid — uses sample)
5. Hierarchical clustering dendrogram (Ward linkage, 300 obs)
6. Agglomerative clustering comparison (k=2–10)

Silhouette scores by k (full training matrix):

  k=2   ████████░░░░░░░░░░░░  0.198  smallest cluster: 6,830 (28.9%)
  k=3   █████████░░░░░░░░░░░  0.221  smallest cluster: 2,100 (8.9%)
  k=4   ████████████████░░░░  0.312  ← strong BUT largest=72% of data
  k=5   ██████████░░░░░░░░░░  0.250  smallest: 526 (unstable)
  k=6   ████████████░░░░░░░░  0.290  ← SELECTED ★  smallest: 583 (2.5%)
  k=7   ████████████░░░░░░░░  0.286  singleton cluster appeared
  k=8+                               singleton clusters appeared

Why NOT k=10 (highest score): singleton cluster (1 observation)
Why NOT k=4 (strong score):   largest cluster = 72% — not meaningful separation
Why k=6: no singletons, stable sizes, interpretable profiles, silhouette 0.290

Cluster profiles at k=6 (training set n=23,657):

Cluster	Label	Size	Share	Key Signal
0	Manager-focused	4,571	19%	is_manager_role=1.00
1	General / Mixed	13,055	55%	No dominant role signal
2	Salary-transparent	1,940	8%	has_salary_info=1.00
3	Data roles	1,451	6%	is_data_role=1.00
4	Software roles	2,057	9%	is_software_role=1.00
5	Entry / low salary	583	2%	Smallest cluster

Official final silhouette score: 0.290 (full training matrix)

Cluster labels one-hot encoded as 6 dummy features. Including clusters improved both regression RMSE and classification F1 over models without them.

---

📈 Regression — Predicting log1p(views)

Baseline

Mean Baseline (predict training mean for all observations):
  RMSE_log = 0.8708   R² = -0.0002   ← floor every model must beat
  MAE_views ≈ 10.64

Baseline Linear Regression (20 features, no clustering):
  RMSE_log = 0.8425   R² = 0.0639

Full Model Comparison (after feature engineering + clustering)

Model	RMSE_log ↓	R² ↑	Notes
Random Forest (Tuned) ★	0.8347	0.0811	RandomizedSearchCV winner
Random Forest (Controlled)	0.8349	0.0807	Manual constraints
Gradient Boosting	0.8370	0.0770	—
Linear Regression + Features	0.8420	0.0640	—
RidgeCV	0.8420	0.0640	—
Lasso Regression	0.8430	0.0640	—
PCA + Linear Regression	0.8440	0.0600	15 components, 96.3% variance
Mean Baseline	0.8708	-0.0002	Floor

Key lessons:

• Unrestricted RF → train R²=0.854, test R²=0.003 (massive overfit). Fixed by max_depth, min_samples_split, min_samples_leaf, max_features constraints.
• 3-fold CV mean RMSE_log: 0.8747 (±0.0125) — stable across folds
• Outlier robustness test: capping views at 99th pct → RMSE_log 0.8147, R²=0.0812

Top Feature Importances (RF Tuned)

description_density          ████████████  #1 — content quality proxy
description_length           ██████████░░  #2 — raw description size
description_word_count       ████████░░░░  #3 — word count
title-description interaction████████░░░░  #4 — combined text signal
is_software_role             ██████░░░░░░  #5 — tech role demand
is_data_role                 █████░░░░░░░  #6 — data role demand
salary_log / has_salary_info ████░░░░░░░░  #7+ — salary signals

desc_salary_interaction ranks #2 in SHAP analysis but further down in Gini importance — both agree on description quality and salary as top drivers.

Why R² = 0.081 Is Acceptable

R² = 0.081 → model explains ~8% of variance in log(views+1)

✓ Beats mean baseline (R²≈0) — real posting-level signal captured
✓ Social engagement inherently noisy — platform factors dominate
✓ 92% of variance from unobservable sources (algorithm, followers, ads)
✓ Practical use = ranking postings, not forecasting exact counts

---

🟠 Classification — High Engagement vs. Normal

Target: high_engagement = 1 if views ≥ 75th percentile of TRAINING views
Class balance: ~75% Normal (Class 0) / ~25% High Engagement (Class 1)
Feature matrix: X_clf uses 24 features (see notebook cell 207)
Metric: F1-score for Class 1 (accuracy misleading with 75/25 imbalance)

Model Comparison

Model	F1 (Class 1)	Recall (Class 1)	Notes
Decision Tree ★	HIGHEST	HIGHEST	max_depth=8, class_weight="balanced"
Logistic Regression	near-best	high	Close to DT in F1
Random Forest	moderate	lower	Lowest FP count
Dummy Baseline	0.00	0.00	Always predicts Class 0

5-fold CV F1: 0.4424 ± 0.0152 — stable, no lucky split

Error Cost Analysis

FN (missed high-engagement) = most costly error
  → Company fails to prioritize, promote, or learn from a strong posting

FP (false alarm) = also costly
  → Recruiter wastes time and budget on a posting that won't perform

Decision Tree minimises FN (catches most high-engagement postings) but produces more FP. Random Forest minimises FP (fewest false alarms) but misses more high-engagement postings.

---

💡 Business Insights

1. Salary transparency is the single highest-leverage action — 74.3% more views for free. Fewer than 30% of postings disclose salary today.
2. Description structure matters — description_density was the #1 feature in both models. Sweet spot: 250–500 words.
3. Tech roles attract disproportionate engagement — is_software_role and is_data_role carry real signal beyond salary.
4. Work type is associated with engagement — contract roles lead, but full-time dominates volume (80%).
5. Platform factors dominate — R²≈0.08 is expected and acceptable. Model value is in ranking postings, not exact prediction.

---

🎁 Bonus Work

🧠 SHAP Explainability

SHAP mean |value| — RF Tuned regression (test observations)

description_density      ████████████  strongest positive impact ↑
desc_salary_interaction  ██████████░░  salary × description synergy ↑
salary_log               ████████░░░░  salary level ↑
has_salary_info          ██████░░░░░░  disclosed → more views ↑
posting_weekend          ██░░░░░░░░░░  weekend → fewer views ↓

desc_salary_interaction ranks #2 in SHAP but lower in Gini — confirms it captures genuine non-linear interaction that neither feature achieves alone.

📊 Feature Importance: Regression vs Classification

                        Regression RF    Classification DT
description_density     #1               #2
desc_salary_interaction #2 (SHAP)        varies
salary_log              #7+              varies
is_entry_role           lower            rises in classification
is_data_role            #6               varies
──────────────────────────────────────────────────────────
Agreement:  description quality + salary dominate both models
Divergence: seniority/role flags matter more for threshold-crossing
            (classification) than for predicting exact counts (regression)

🔬 Additional Extras

• Interactive K-Means Widget — explore different k values visually (notebook cell 4.11)
• Hierarchical Clustering Dendrogram — Ward linkage, 300 obs sample (cell 4.12)
• Agglomerative Clustering Diagnostic — k=2–10 comparison (cell 4.13)
• Outlier Robustness Test — views capped at 99th percentile: RMSE_log 0.8147 vs 0.8347 uncapped
• 3-fold CV for regression — mean RMSE_log 0.8747 ± 0.0125

---

🛠️ How to Use the Models

import pickle, numpy as np

with open("linkedin_regression_model.pkl", "rb") as f:
    reg_model = pickle.load(f)
with open("linkedin_classification_model.pkl", "rb") as f:
    clf_model = pickle.load(f)

# Regression — predict log(views+1), convert back to raw view estimate
log_views_pred = reg_model.predict(X_test_fe)
views_pred = np.expm1(log_views_pred)

# Classification — predict high-engagement label (0 = Normal, 1 = High)
label = clf_model.predict(X_clf)

Regression model expects 30-column X_test_fe (including cluster dummies). Classification model expects 24-column X_clf (see notebook cell 207). Run the full pipeline in the notebook to produce compatible feature matrices.

---

Assignment 2 — Classification, Regression, Clustering, Evaluation | LinkedIn Job Postings · arshkon/linkedin-job-postings (Kaggle)