Automatically matches prospective tenants and property owners.
Each post contains the location, property type and price
Determine if a post is from a property owner or a prospective tenant
Extract the location, property type and price from the post
- If the post is from a service provider then the location is a single point in an area
- If the post is from a prospective tenant the location may be:
- A single area,
- A list of areas,
- An area and a range (Looking for a place around Montshioa, looking for a place near Mega City)
- Commune,
- Room,
- Bachelor(a room with a kitchenette and in suite bathroom),
- 1 bedroom,
- 2 bedroom,
- 3 bedroom,
- 4 bedroom
- If the post is from a service provider then the price is a fixed value,
- If the post is from a prospective tenant then the price is a range.
- Train intent recognition model,
- Train named entity recognition model,
- Service
- Determine intent,
- Extract features using NER,
- IF Prospective tenant:
- Search Supplier DB
- IF Supplier not found: Add request to Prospective Tenant DB
- ELSE: Return supplier list
- IF Property owner:
- Add supplier to Supplier DB
- Python,
- Rasa NLU,
- BERT