IntroductiontotheWeaviateSmartGraph
@bobvanluijt
What’stheproblem?
3
4
SELECT * FROM `actors` WHERE `name` = ‘Brad Pitt’
5
Brad Pitt ?
is a(n)
6WikiMedia
7Getty Images
8
A Bain & Company survey of 325 global companies found that;
1. 59% believe they lack the capabilities to generate
meaningful business insights from their data.
2. 85% said they will require substantial investments
to update their data platform.
What’sthesolution?
10
Smart Graphs
11
A [smart]graph acquires and integrates
information into an ontology and applies a
reasoner to derive new knowledge.
“
Towards a Definition of Knowledge Graphs
12
Howtocreateyourown?
14
15
Weaviate Smart Graph
16
Apps
etcetera
UseCaseFeatures
18
Semantic Search
19
Automatic
Classification
20
Knowledge
Representation
DeveloperFeatures
22
Contextionary
P2P Network
GraphQL Containerized
ScalableHTTP API
Weaviate’sContextionary
24
25
company
apple
fruit
[0.34534, 2.3543, 8.34532 … ]
600 dimensions
stanley kubrick
2001 a space odyssey
26
Company, name, Apple,
founded, in, 1976, in,
country, with, the, name,
United, States
Company:
name: Apple
foundedIn: 1976
inCountry:
"$cref": ...
hasCeo:
Person:
name: Tim Cook
27
Company, name, Apple,
founded, in, 1976, in,
country, with, the, name,
United, States
[0.34534, 2.3543, 8.34532 … ]
28
29
company
iphone
30
{
Get{
Things{
Company(explore:{
concepts: ["fashion"]
}){
name
}
}
}
}
31
{
"data": {
"Local": {
"Get": {
"Things": {
"Company": [
{
"name": "Vogue"
}
]
}
}
}
},
"errors": null
}
Demo
Usecaseexamples
Retail & Consumer Goods
34
Description: Find customer white spots in public
datasets.
Triggers: manual search in cities.
Business case: new customers are found manually,
Weaviate automates this which saves money.
USPs used: knowledge representation and semantic
search.
Cyber Security
35
Description: Match cybersecurity notifications to
threads.
Triggers: Cybersecurity threads are often matched
manually.
Business case: Cybersecurity threads are often
matched manually, Weaviate automates this which
is faster and cheaper.
USPs used: Automatic classification
Oil & Gas
36
Description: relate thousands of databases.
Triggers: different databases uses different
language to describe the same thing.
Business case: automatically find the location of
data without manual intervention.
USPs used: automatic classification and knowledge
representation.
www.semi.technology
Thankyou!
www.semi.technology
bob@semi.technology
@bobvanluijt

Weaviate Smart Graph - Bob van Luijt (SeMI Technologies)