What Finance can learn from Dating Sites
23 likes4,151 views
The document discusses how insights from dating sites can inform financial systems through the analysis of five key graphs: payment, customer, entitlement, asset, and master data graphs. It emphasizes the importance of relationships and data connectivity for real-time query performance, proposing the use of Neo4j as a solution for managing complex data relationships effectively. The text also contrasts traditional relational databases and NoSQL databases with Neo4j, highlighting its advantages in speed and scalability.
![MATCH
(:Person
{
name:“Dan”}
)
-‐[:FRIENDS]-‐>
(:Person
{
name:“Ann”}
)
FRIENDS
Dan Ann
Label Property Label Property
Node Node
Cypher
Query
Language](https://image.slidesharecdn.com/whatcanfinancelearnfromdatingsites-150512063302-lva1-app6891/85/What-Finance-can-learn-from-Dating-Sites-14-320.jpg)
![MATCH
(boss)-‐[:MANAGES*0..3]-‐>(sub),
(sub)-‐[:MANAGES*1..3]-‐>(report)
WHERE
boss.name
=
“John
Doe”
RETURN
sub.name
AS
Subordinate,
count(report)
AS
Total
Express
Complex
Queries
Easily
with
Cypher
Find
all
direct
reports
and
how
many
people
they
manage,
up
to
3
levels
down
Cypher
QuerySQL
Query](https://image.slidesharecdn.com/whatcanfinancelearnfromdatingsites-150512063302-lva1-app6891/85/What-Finance-can-learn-from-Dating-Sites-15-320.jpg)
![Express
Complex
Relationship
Queries
Easily
For
a
given
fund,
return
all
assets
that
are
made
up
of
other
assets,
ordered
by
the
total
number
of
included
assets
Cypher
Query
SQL
Query
MATCH
(fund)-‐[:INCLUDES*0..n]-‐>(sub),
(sub)-‐[:INCLUDES*1..n]-‐>(asset)
WHERE
fund.ticker
=
“TRLGX”
RETURN
sub.ticker
AS
Asset_Group,
count(asset)
AS
Total
ORDER
BY
Total
DESC](https://image.slidesharecdn.com/whatcanfinancelearnfromdatingsites-150512063302-lva1-app6891/85/What-Finance-can-learn-from-Dating-Sites-60-320.jpg)
![Cypher
Query:
Movie
Recommendation
MATCH
(watched:Movie
{title:"Toy
Story”})
<-‐[r1:RATED]-‐
()
-‐[r2:RATED]-‐>
(unseen:Movie)
WHERE
r1.rating
>
7
AND
r2.rating
>
7
AND
watched.genres
=
unseen.genres
AND
NOT(
(:Person
{username:”maxdemarzi"})
-‐[:RATED|WATCHED]-‐>
(unseen)
)
RETURN
unseen.title,
COUNT(*)
ORDER
BY
COUNT(*)
DESC
LIMIT
25
What
are
the
Top
25
Movies
• that
I
haven't
seen
• with
the
same
genres
as
Toy
Story
• given
high
ratings
• by
people
who
liked
Toy
Story](https://image.slidesharecdn.com/whatcanfinancelearnfromdatingsites-150512063302-lva1-app6891/85/What-Finance-can-learn-from-Dating-Sites-71-320.jpg)
What Finance can learn from Dating Sites
- 1. What Finance can learn from Dating Sites Max De Marzi GOTO Chicago
- 2. About Me • Max De Marzi -‐ Neo4j Field Engineer • My Blog: http://maxdemarzi.com • Find me on Twitter: @maxdemarzi • Email me: maxdemarzi@gmail.com • GitHub: http://github.com/maxdemarzi
- 3. TLDR:
- 4. http://www.gartner.com/id=2081316 Consumer Web Giants Depend on Five Graphs Social Graph Mobile Graph Intent Graph Interest Graph Payment Graph
- 5. Friends of Friends Graph • Real World Basis • Its Weighted (BFF vs Family) • Awesome or Awkward The Five Graphs of Love Location Graph • Long Distance Sucks • Where are the single people • Where should we meet Passion Graph • Shared Interests • Desired Traits • Long vs Short Term Safety Graph • True Identity • Liers and Cheaters • Balancing Privacy SPAM Graph • Click Bait • Wanna Cam? • Professionals 1 2 3 4 5
- 6. The Five Graphs of Love 1Friends 2Passion 3Location 4Safety 5SPAM
- 7. Meet Jeremy • Single • Handsome • Friendly • etc Jeremy
- 9. Bang With Friends => Down
- 10. Better Idea
- 12. Deeper
- 13. Friends of Friends of Friends FOFOFs
- 14. MATCH (:Person { name:“Dan”} ) -‐[:FRIENDS]-‐> (:Person { name:“Ann”} ) FRIENDS Dan Ann Label Property Label Property Node Node Cypher Query Language
- 15. MATCH (boss)-‐[:MANAGES*0..3]-‐>(sub), (sub)-‐[:MANAGES*1..3]-‐>(report) WHERE boss.name = “John Doe” RETURN sub.name AS Subordinate, count(report) AS Total Express Complex Queries Easily with Cypher Find all direct reports and how many people they manage, up to 3 levels down Cypher QuerySQL Query
- 16. The Five Graphs of Love 1Friends 2Passion 3Location 4Safety 5SPAM 2Passion Graph
- 18. Likes
- 19. Traits
- 20. The Five Graphs of Love 1Friends 2Passion 3Location 4Safety 5SPAM 3Location
- 21. Location
- 22. Three Dots and a Dash
- 23. Recommend Love Find your soulmate in the graph • Are they energetic? • Do they like dogs? • Have a good sense of humor? • Neat and tidy, but not crazy about it? What are the Top 10 Potential Mates for me • that are in the same location • are sexually compatible • have traits I want • want traits I have
- 24. Cypher Query: Love Recommendation
- 26. The Five Graphs of Love 1Friends 2Passion 3Location 4Safety 5SPAM 4Safety
- 28. Liars
- 30. Let’s take a closer look at Jonathan Jonathan
- 31. Follows
- 32. Tweets
- 34. The Five Graphs of Love 1Friends 2Passion 3Location 4Safety 5SPAM 5SPAM
- 35. Click Bait
- 36. Cam Girls/Boys
- 37. Professionals
- 38. Payment Graph • Fraud detection • Credit risk analysis • Chargebacks Financial Giants Depend on Five Graphs As Well Asset Graph • Portfolio analytics • Risk management • Market/sentiment analysis • Compliance Customer Graph • Org drill-‐through • Product recommendations • Mobile payments Entitlement Graph • Identity management • Access management • Authorization Master Data Graph • Enterprise collaboration • Corporate hierarchy • Data governance 1 2 3 4 5
- 39. The Five Graphs of Finance 1Payment 2Customer 3Entitlement 4Asset 5Master Data 1Payment Graph
- 40. Intuit Payment Graph Discover latent network from multiple product data stores • Uniquely identify entities and their connections • Connections scored by volume of trade Empower business-‐unit teams to leverage the Intuit Payment Graph to build applications • Graph to be available for real-‐time queries 1Payment
- 41. Consumer Profile Facets Identity Name Address Phone Email Social Facebook Yelp Twitter … Demographics Age Gender … Business Profile Facets Identity Name Address Phone Email Social Facebook Yelp Twitter … Demographics Category Revenue Employees … 1Payment Payment Graph Depends on the Customer Graph
- 42. 1Payment Capturing C2B and B2B Transactions BUSINESSBUSINESS CONSUMER June 1 purchase $25.95 June 3 purchases $650.25
- 43. PRODUCT Name: Zeta … PRODUCT Name: Payroll … COMPANY Name: Viva LLC Zip: 94040 … COMPANY Name: Beta LLC Zip: 94043 … COMPANY Name: Acme, Inc. Zip: 95134 … Relationship CUSTOMER Transactions: 467 Years: 3Relationship LICENSE Years: 8 Relationship CUSTOMER Transactions: 125 Years: 1 Relationship LICENSE Years: 3 #1: Payment Graph Example
- 44. Streamlining credit card chargebacks Cardholder calls card issuer to dispute transaction Cardholder receives credit card statement Card issuer returns transaction through card network Acquirer resolves chargeback or forwards it to merchant Merchant receives chargeback and accepts or challenges it Acquirer forwards representation to card network Card issuer verifies representation and credits cardholder Network verifies and forwards representation to card issuer #1: Payment Graph Example 1Payment
- 45. 1Payment 2Customer 3Entitlement 4Asset 5Master Data The Five Graphs of Finance 2Customer Graph
- 46. All Companies and Customers Are Related 2Customer
- 47. The Corporate Hierarchy is really a graph 2Customer Corporate Hierarchy is Really a Graph
- 48. Name Windsor Press, Inc. Address 6 North Third St City Hamburg State PA Zip 19526 Phone 610-‐562-‐2267 Name The Windsor Press Address 6 North 3rd Street City Hamburg State PA Zip 19526-‐0465 Phone 610-‐562-‐2267 ID 002114902 Name Windsor Press, Inc. Address 6 N 3rd St City Hamburg State PA Zip 19526-‐1502 Phone 610-‐562-‐2267 Both of the records above map to the same record 2Customer Cleaning and Matching for 360-‐Degree Master View
- 49. Synthetic Identities and Fraud Rings 145 Hickory Rd Pasadena, CA 415 Hickory St Pasadena, CA 626-‐407-‐1234 626-‐814-‐6532 Quickly see which customers share the same contact information 2Customer
- 50. 3 fake addresses and 3 fake phone addresses can create 9 fake customers 2Customer Bank Fraud Using False Personas
- 51. High Speed Fraud -‐ 1000 R/S http://maxdemarzi.com/2014/02/12/online-‐payment-‐risk-‐management-‐with-‐neo4j/
- 52. High Speed Fraud -‐ 8000 R/S http://maxdemarzi.com/2014/02/27/neo4j-‐at-‐ludicrous-‐speed/
- 53. High Speed Fraud -‐ 28000 R/S http://maxdemarzi.com/2014/03/10/its-‐over-‐9000-‐neo4j-‐on-‐websockets/
- 54. 1Payment 2Customer 3Entitlement 4Asset 5Master Data The Five Graphs of Finance 3Entitlement Graph
- 55. http://blogs.gartner.com/ian-‐glazer/2013/02/08/killing-‐iam-‐in-‐order-‐to-‐save-‐it/ Killing IAM in Order to Save It 3Entitlement
- 56. Access Control
- 58. 1Payment 2Customer 3Entitlement 4Asset 5Master Data The Five Graphs of Finance 4Asset Graph
- 59. Portfolio Analytics Asset Graph Examples IT Asset Management Risk Analysis 4Asset
- 60. Express Complex Relationship Queries Easily For a given fund, return all assets that are made up of other assets, ordered by the total number of included assets Cypher Query SQL Query MATCH (fund)-‐[:INCLUDES*0..n]-‐>(sub), (sub)-‐[:INCLUDES*1..n]-‐>(asset) WHERE fund.ticker = “TRLGX” RETURN sub.ticker AS Asset_Group, count(asset) AS Total ORDER BY Total DESC
- 61. The Five Graphs of Finance 1Payment 2Customer 3Entitlement 4Asset 5Master Data 5Master Data Graph
- 65. Linked Data Connect to the Semantic Web
- 66. Technology for Managing the Five Graphs of Finance 1Payment 2Customer 3Entitlement 4Asset 5Master Data
- 67. Intercontinental Exchange Social network for brokers 5Master Data
- 70. Movie Data Model
- 71. Cypher Query: Movie Recommendation MATCH (watched:Movie {title:"Toy Story”}) <-‐[r1:RATED]-‐ () -‐[r2:RATED]-‐> (unseen:Movie) WHERE r1.rating > 7 AND r2.rating > 7 AND watched.genres = unseen.genres AND NOT( (:Person {username:”maxdemarzi"}) -‐[:RATED|WATCHED]-‐> (unseen) ) RETURN unseen.title, COUNT(*) ORDER BY COUNT(*) DESC LIMIT 25 What are the Top 25 Movies • that I haven't seen • with the same genres as Toy Story • given high ratings • by people who liked Toy Story
- 72. Relational Databases Can’t Handle Relationships Well • Cannot model or store data and relationships without complexity • Performance degrades with number & levels of relationships, and database size • Query complexity grows with need for JOINs • Adding new types of data and relationships requires schema redesign, increasing time to market … making traditional databases inappropriate when relationships are valuable in real-‐time Slow development Poor performance Low scalability Hard to maintain
- 73. NoSQL Databases Don’t Handle Relationships • No data structures to model or store relationships • No query constructs to support relationships • Relating data requires “JOIN logic” in the application • No ACID support for transactions … making NoSQL databases inappropriate when relationships are valuable in real-‐time
- 74. Real-‐Time Query Performance Performance must hold steady with scale Connectedness and Size of Data Set Response Time 0 to 2 hops 0 to 3 degrees Thousands of connections Tens to hundreds of hops Thousands of degrees Billions of connections Relational and Other NoSQL Databases Neo4j Neo4j is 1000x faster Reduces minutes to milliseconds
- 75. Re-‐Imagine Your Data as a Graph Neo4j is an enterprise-‐grade graph database that enables you to: • Model and store your data as a graph • Query relationships with ease and in real-‐time • Seamlessly evolve applications to support new requirements by adding new kinds of data and relationships Agile development High performance Vertical and horizontal scale Seamless evolution
- 76. THANK YOU