Can Federation and Semantic Technologies Get Us Closer to the Answers We Seek on the Web?

schoolmistInternet and Web Development

Oct 22, 2013 (3 years and 10 months ago)

80 views

Can Federation and

Semantic Technologies

Get Us Closer to the Answers

We Seek on the Web?






Colin Britton
-

Metatomix

Gilbane Boston
-

November 2007

Introduction


Colin Britton


CTO and Co
-
founder Metatomix


Dedham, MA Headquarted


Previous Experience


SVP Technology MediaBridge


Digital asset management and Content management for print


Digital Media background


5 Patents in Enterprise Semantic Technology

What do we mean by Semantic Technology?


Solutions that are based around the W3C standards


RDF

Resource Description Framework


OWL


Web Ontology Language


SPARQL


Semantic Query Language



Solutions that transform


Data to Information


Information to Knowledge


Knowledge to Action


Today’s Business Challenges

“80% of the effort in delivering any Risk
-
based project is spent on data management;
identifying the data sources, getting access
to them, integrating the information and
understanding it.”

Ray O’Brien

Global Head of Risk Technology, HSBC Investment Bank

Consider this:


A siloed approach to customer data
collection has created the challenge of
resolving semantic mismatches
between systems to gain a 360 view of
a customer.

Enterprise Data Challenges


The Reality


BI architecture complexity has lead to fragmented data and disaggregate view
of metrics.


It does not help the cause that these metrics are dispersed over many BI tools,
many corporate data sources, and many competitive information sources.



Explosion in data volumes dispersed over many core applications like CRM,
ERP, SCM, Hr, etc.


With disparate and fragmented systems it can be quiet a challenge to get a
‘unified view of enterprise metrics' .


Most organizations use point based solutions / tools for analysis and use
traditional methods of phone / email to drive actions.


Getting timely, in
-
context information to make critical decisions.


For effective decision making it is important real time data is available for
collaboration.



Semantic Technology Approach

Integrate your Application
Specific Data Silos

Semantic Rules Engine to
apply a common
understanding across
applications

Embrace the Distributed
Enterprise Landscape

SWT
simplifies the way in which an application uses data from ANY data source

Result: A
real
-
time

virtual integration layer that unifies the applications across your
enterprise landscape and allows you to see relationships amid 1000’s of seemingly
unrelated events

Strategic Investigation

Business

Process

Data

Business

Rules

Strategic Investigation

Business

Process

Data

Business

Rules

Knowledge

Void

Fully
Integrated
Suite of
Capabilities
all in a

Single
Platform

Current Technology Limitations


Structured, semi
-
structured and unreliably structured
data needs to be brought together


Traditional techniques are static or too simplistic


Common understanding needs establishing


Most technology bind the understanding to the storage or logic
systems

such as the RDBMS Schema


Links in data need to be created


Business Processes are brittle with variable data


The traditional waterfall BPM or workflow generates large
amounts of exceptions when given inconsistent data


COMMON MODEL

Model your Business

Our Semantic Services uses a common business model to explicitly define both an organization’s
processes and relationships in rich detail, abstracted from individual systems and data


Built around business process


Understands structured &
unstructured data


Model can be added easily
without having to change the
systems they refer to.


Cost Effective


Real time…no out of
sync


Easy to manage across
organizational
boundaries

Correlating data within
the model creates

new
knowledge

and enables
strategic
investigation

Common Model + Rules = Real Time Actions


+

+

Rules

Integrated development
and runtime environment


Organizational knowledge and
policies are explicitly defined.


Differences in meaning and
context between data reconciled.


New knowledge identified.


Rules engine consults the model
for understanding and polices,
dynamically evaluating facts
against policies to make
decisions.


Manages the applications, gathers
data non
-
invasively and executes
actions based on decisions from the
rules engine

A patented combination of model based integration and dynamic rules within an integrated development

and runtime environment, creating
integration solutions that think and act in real
-
time based on
organizational policies.

COMMON MODEL

Ontology

Editor

Repositories (Workspace, Model, Metadata)

UI (Flex and Struts)

Data Providers / Consumers

Model Manager

Rules Engine

Runtime & Process Chains

DP Editor

RDF Editor

Rule Editor

Admin App

FlexBuilder


Ontologies


Rules

Templates

Instance

Data

SWT Event Flow

Integrated Justice Applications

The Process

Automating, Aligning and Accelerating

Incident

Investigation

Arrest
Made

Warrant
Issued

Warrant
Requested

First
Appearance

Court Case
is Heard

Sentencing

Incarceration

Release

Probation

Judicial Process

Corrections

Law Enforcement

Judicial Inquiry System
(JIS)


Connects all players of the judicial
process and leverages the
“universe” of data sources
enabling real
-
time, intelligent
sharing of information across
local, state, and federal systems
from a single interface

Judicial Data Exchange

(JDX)


Provides real
-
time information exchange between any stakeholders in the judicial process to automate
manual processes, eliminate duplicate entry, improve data quality and increase the efficiency and
accuracy during the judicial process
(i.e.


warrant processing, indictments, bond notification, arraignments, etc.)

Criminal Investigation
(CI)


Enables investigators to
quickly and easily query
multiple data sources across
multiple agencies to
investigate individuals and
crimes

Judicial First
Appearance
(JFA)

Correlates criminal and civil
history files from various
agency systems, evaluates
threat level using pre
-
defined
rules and sends notification
(i.e.


Jessica Lunsford Act)

Metatomix Suite of Integrated Justice Applications

Judicial Inquiry System (JIS)


Clerks manually had to
look up pieces of
information in over a
dozen systems, across
multiple agencies


Different sign
-
ons ,
different queries, different
data formats


Clerks had to manually
piece together
background data


Average 45 minutes per
background check


Inconsistent, inaccurate
results

State’s Attorney’s

Dept. of Corrections

The Courts

Law Enforcement

NCIC

Federal Systems

Sentencing

Appriss

(Local Jails)

Judge

Calendering

Local Systems / Agencies

Records

Mgmt

Computer

Aided

Dispatch

Case Mgmt

AFIS

Clerk Data

DMV

FDLE

DOC

DJJ

Sex Offender

Registry

Dept of Children

& Family

State

Judicial Inquiry System


Consolidated Subject Profile


Alias Management


Digital image and document viewing


Risk assessment/Automatic notification


Judge calendaring management

Flex


Simple interface and single query


Web Services (SOAP, REST)

System deployed for less then 10% of the cost

and 10x quicker than custom integration estimates

Average time to perform background
check decreased from 45 minutes to 3.

Used by approx 5,000 users daily


Using Justice Standards for process
frameworks (JIEM) and data standards


Data Services


GJXDM



Department of Motor Vehicles



State office of Homeland Security



State Bureaus of Investigation



Drugs and Narcotics Agency



Department of Natural Resources



National Crime Information Center




Vital Statistics Bureau



Children/Family services



Corrections



Probation



Department of Revenue



Ports Authority


Web based single sign
-
on decreased
interfaces from over 12 to 1

Real
-
time, secure access to multi
-
agency
data. No changes to existing systems and
no need for data warehouses.

JIS Solution Barriers Removed


Disparate data sources


Organizationally


Politically


Technologically


Shifting Data


System changes


Ownership limitations


Accessibility




Airbus: Real Time Costing

Airbus: Real
-
Time Costing Visibility

Current Cost
Knowledge

Product

Specification

Cost Definition

FEASIBILITY

CONCEPT

DEFINITION

DEVELOPMENT

POST EIS

M0

M3

M4

M5

M6

M7

M13

M14

*Not to Scale

T A R G E T S E T T I N G

VISION

To enable Cost Awareness at M0 & cost knowledge



to grow during the development process

20

21

Step 6
-

Cost Impact analysis

Summary display of cost
and other key data points
for model originally
predicted from Tadpole
definition or as revised by
MIMES data

Airbus


Real
-
time costing visibility

Client calls Semantic Services via SOAP

Existing legacy systems and processes
being invoked across organizational and
technological boundaries

Data via SOAP, Flat file and JDBC

Single view of cost across organization and
product lifecycle. Clearly identify accurate
costing earlier to influence design changes.
Understand impact of design changes on
cost as fast as they occur.


Defining one common aircraft definition


Correlate with manufacturing data


Compute costing from financial systems


Design studies on impact and cost of change


Iterative monitoring for changes

Flex

Web Services (SOAP, REST)


Data Services

SAP

CAD/CAM

Simulation

Models

Parametric

Models

Design to Cost



FInance

MFTG

AERO

Product

Mgmt

Airbus Solution Barriers Removed


Complex design process


60+ different tools


Long development lifecycle


Human knowledge based


Complex parametric analysis


Simplistic Sharing


Spreadsheet data sharing


Report delivered in PowerPoint


Meeting centric knowledge sharing

Semantic Web Technology Summary


Ideal for mixed data information sharing


Change ready solutions


change major impact on solution cost


Declarative technology approach


More in tune with how humans process


Enterprise ready


Leverages the assets in
-
place