The document discusses the benefits of connecting previously unconnected data and institutional knowledge. It states that siloed and inefficient analytics can lead to increased costs and risks due to duplication and scattered decision-making. An integrated approach enabled by tools like the Parabole TRAIN platform can help extract knowledge graphs from diverse data sources, allow for more informed decisions, and make organizations more adaptable to future changes.
Unconnected data risks and siloed operations costs
1. Unconnected data is a liability
Institutional knowledge is an asset
COSTS AND RISKS OF INEFFICIENT ANALYTICS SOARS
SILOED OPERATIONS HAVE LED TO DUPLICATION
DECISION-MAKING MORE COSTLY AND POROUS DUE TO SCATTERED DOMAIN
PERSPECTIVES
ADAPTIVE TO FUTURE CHANGES
ENABLES INFORMED DECISION-MAKING
1
2. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
2
3. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
3
4. • Generate Industry – Territory Risk Analysis
• 50 states of USA are considered as the Territory for analysis
• 11 Industry types (as classified by SASB Materiality Map) are considered for analysis
• Industry need data are collected from different industry document including SASB Materiality Map
• Territory data on Crime, Water Stress, Climate, Bio-diversity zone, Native land zone are gathered from different
sources
• Parabole TRAIN platform is used for extracting structure knowledge graph automatically from Industry and
Territory data (Unstructured, Semi-structured)
• Cambridge Semantics AnzoGraph DB is used for storage, inference and analysis of Knowledge Graph
Machine Teaching (TRAIN): Use Case Under Analysis
4
Use Case Objective
5. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
5
TerritoryRiskProfile
Natural Disaster
Bio-diversity Conflict
Native Land Conflict
6. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
6
A natural disaster is a major adverse
event resulting from natural
processes of the Earth; examples
are floods, hurricanes, tornadoes, vol
canic
eruptions, earthquakes, tsunamis, st
orms, and other geologic processes.
A natural disaster can cause loss of
life or damage property,[1] and
typically leaves some economic
damage in its wake, the severity of
which depends on the affected
population's resilience (ability to
recover) and also on the
infrastructure available.
7. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
7
8. Machine Teaching (TRAIN): SME Knowledge, Data Ingestion
TRAIN
Institutional Assets,
SME knowledge
General Assets
Cognitive models
Knowledge graph
Inst. taxonomy
8
14. Machine Teaching (TRAIN): Knowledge Graph Generation
Industry
Mining
Natural
Disaster
High
Territory
Texas
Natural
Disaster
Flood
Tornado
15. Knowledge Graph Storage and Inference
Industry
Mining
Natural
Disaster
High
Territory
Texas
Natural
Disaster
Flood
Tornado
16. Graph Database Capabilities
Scalability
Inferencing with RDFS+
Get more from the relationships
in your data with inferencing
Data Science
Use the included data science
algorithms for data prep and other
critical functions
Languages
Graph Algorithms
PageRank , shortest path and other graph
algorithms are made possible with RDF*
and SPARQL*
Analytics
Some capabilities to consider when performing analytics from graph
Is the graph database able to use multiple servers to
load and analyze data for ultimate performance?
Geospatial
Massively Parallel
Custom Algorithms
Write your own customizations in
JAVA or C++ with AnzoGraph’s SDK