4. Recommendations apply to many industries
• Retail & eCommerce
• Products
• Services
• Media & Content
• Events & Activities
• Social & Dating
• Education
• Travel
• Real Estate
5.
6. Goals of a good Recommender System
• Relevance
• Users are more likely to consume Items they find interesting
• Novelty
• Items they didn’t see/buy/view in the past (related)
• Serendipity
• Unexpected Items, a lucky discovery (unrelated)
• Increasing recommendation diversity
• Same type of items increases the risk of the User not liking any Item
14. Silos & Polyglot Solutions
Purchases
RELATIONAL
DB
Product
Catalogue
DOCUMENT
STORE
WIDE COLUMN
STORE
Views
DOCUMENT
STORE
User Review
RELATIONAL
DB
In-Store
Purchase
Shopping
Cart
KEY VALUE
STORE
Good for Analytics, BI, Map
Reduce
Non-Operational, Slow
Queries
15. Silos & Polyglot Solutions
Purchases
RELATIONAL
DB
Product
Catalogue
DOCUMENT
STORE
WIDE COLUMN
STORE
Views
DOCUMENT
STORE
User Review
RELATIONAL
DB
In-Store
Purchase
Shopping
Cart
KEY VALUE
STORE
Connector
Real-Time
Queries
21. Accuracy & Control
• Content filtering
• Collaborative filtering
• Expert/Supervised Intelligence
• Cognitive/Machine Intelligence Easily Plugged In
• Feedback from consumers
• Model transparency
22. Agile & Flexible
• Evolve & tune model and recommendations organically, in response
to changing business requirements, with no downtime
• Integrate with cognitive/machine learning and NLP environment
• Scoring features can be easily extended
• Create & tune rules and incorporate feedback "on the fly"
• Deploy quickly
• Recommendation Engine can be reverse engineered from existing
data sources