Rent Partner Recommendation System

A recommendation system that helps tenants find compatible roommates for co-renting apartments using Neo4j graph database and a weighted compatibility scoring algorithm.

Application Logic

Compatibility Scoring Algorithm

The system calculates compatibility between two tenants using a weighted scoring system:

Factor	Weight	Logic
Cleanliness Match	20 pts	Awarded if cleanliness scores are within 2 points of each other
Schedule Match	20 pts	Awarded if both are night owls OR both are early birds
Work Style Match	15 pts	Awarded if both work from home OR both go to office
Shared Interests	5 pts each	Each common interest adds 5 points
Pet Status Match	10 pts	Awarded if both have pets OR both don't have pets

Minimum Score: 30 points required for a recommendation

Pre-Filtering Rules

Before compatibility scoring, tenants must pass these filters:

• Combined budget ≥ apartment rent
• Matching smoking preference (both smokers OR both non-smokers)
• Pet compatibility (if either has pets, apartment must allow pets)
• Both interested in the same apartment

Data Model

Nodes:

• Tenant: id, name, age, occupation, budget, smoker, hasPets, petType, cleanliness (1-10), nightOwl, workFromHome, interests[], phone, email
• Apartment: id, address, city, bedrooms, bathrooms, rent, deposit, sqft, parking, furnished, petsAllowed, amenities[]
• Landlord: id, name, phone, email, rating

Relationships:

• OWNS: Landlord → Apartment (properties: since)
• INTERESTED_IN: Tenant → Apartment (properties: moveInDate, urgency)

Core Functions

1. Find Roommates for Tenant

Input: tenant_id, apartment_id
Process:
  1. Query all tenants interested in same apartment
  2. Filter by budget, smoking, pet policies
  3. Calculate compatibility scores
  4. Filter by minimum score (30)
  5. Sort by score DESC, then budget surplus DESC
Output: Ranked list of compatible roommates

2. Find All Compatible Pairs

Input: apartment_id
Process:
  1. Query all tenants interested in apartment
  2. Generate all possible pairs (t1, t2) where t1.id < t2.id
  3. Apply filters and compatibility scoring
  4. Filter by minimum score (25)
  5. Sort by score DESC
Output: Ranked list of tenant pairs

3. Calculate Financial Feasibility

Combined Budget = tenant1.budget + tenant2.budget
Budget Surplus = Combined Budget - apartment.rent
Each Tenant Share = apartment.rent / 2

Project Structure

src/
├── db_connection.py           # Neo4j queries and connection
├── recommendation_engine.py   # Compatibility scoring logic
└── main.py                    # Interactive CLI