+ All Categories
Home > Documents > Semantic Web @ BBN Application to the Digital Whitewater Challenge

Semantic Web @ BBN Application to the Digital Whitewater Challenge

Date post: 01-Jan-2016
Category:
Upload: evangeline-marks
View: 25 times
Download: 0 times
Share this document with a friend
Description:
Semantic Web @ BBN Application to the Digital Whitewater Challenge. Mike Dean [email protected] John Hebeler [email protected]. Outline. BBN Introduction The Digital Whitewater Challenge Applications Future Directions. BBN Technologies Overview. - PowerPoint PPT Presentation
16
Copyright ©2009 BBN Technologies Semantic Web @ BBN Application to the Digital Whitewater Challenge Mike Dean [email protected] John Hebeler [email protected]
Transcript

Copyright ©2009 BBN Technologies

Semantic Web @ BBNApplication to the

Digital Whitewater Challenge

Mike [email protected]

John [email protected]

Copyright ©2009 BBN Technologies2

Outline

BBN Introduction

The Digital Whitewater Challenge

Applications

Future Directions

Copyright ©2009 BBN Technologies

BBN Technologies Overview

An advanced technology research and development firm, specializing in Information, Computer, & Physical Sciences

Known for technical excellence and challenging conventions to provide new and fundamentally better solutions to complex technical problems

Providing effective, real-world solutions and satisfying our customers and have been key to our success for over 60 years

Our staff consists of ~800 professionals • 2/3 with advanced degrees

We maintain principal offices in Cambridge, MA and the Washington, DC area

Copyright ©2009 BBN Technologies

History of Innovation

1950s

Acoustic Design for UN General Assembly Hall

AI Pattern Recognition

Program

1960s

1st Time Sharing Demonstration

LOGO Programming

Language

ARPANET-First Multi-node Packet Switched Network

1970s

1st Person-to-Person Network

Email

@ Sign for Email Addresses

Acoustic analysis of JFK

Assassination Tapes

Analysis of Nixon Watergate Tapes

First Symmetric Multi-processor

First TCP for UNIX

KL-ONE Knowledge

Representation System

1980s

1st Electronic Mail

Defense Data Network

National Science Foundation

Network (NSFNET)

Natural Language Computer Interface

Intelligent Agents

SimNet

Collaboration Planning

Technology

Cronus Distributed Computing

Environment

1990s

Secure email for DoD

Multi-Gigabit Router

DARPA Information Assurance

Broadband Wireless Technology

Genetic Algorithm Scheduling Tools

Collaborative Planning for Desert

Storm

ATM Switch

40K Word Speech Recognition System

Logistics Anchor Desk deployed for

Bosnia

Safekeyper Certificate Management

Certification Authority Workstation (CAW)

2000s

Call Director Natural Language

Routing

Semantic Web – OWL, SWRL,

OWL-S

Asio tools forNet-Centric Data

Sharing and Enterprise Web

Services

Microthunder Urban

Environment Surveillance 

System

Cognitive Agent Architecture -

Cougaar

Boomerang Acoustic Shooter Detection System

Quantum Cryptographic

Network

Copyright ©2009 BBN Technologies5

Semantic Web @ BBN

Research– DARPA Agent Markup Language (DAML)

– Geospatial

– Prediction

– Machine Learning

Tools– Parliament

– Snoggle

– Asio ScoutTM

Collaboration– semwebcentral.org

– Peer-reviewed papers, conference sponsorship, and books...

– Working groups

Applications– Broad range of government and commercial applications

from research explorations to production deployment

InnovationLoop

Copyright ©2009 BBN Technologies6

The Digital Whitewater Challenge

Volume– ~ 8 exabytes/month

– AT&T alone – 18 petabytes/day

– youtube alone is now more than all Internet traffic in 2000

Variety– Modes: 40% video

– Formats: IM, Email, Blogs, XML,

– Places: MySpace, Facebook, Second Life, Darkweb...

– Trillion unique URLs

– Fractal – Laptop, Enterprise, WWW

Velocity

– 38 minutes to find one document

– 10 second lead worth millions/lives

BUT...

– Is it true?

– Is it useful?

– Can you find it in time?

Copyright ©2009 BBN Technologies

Semantic Web Application Lifecycle

Model Creation• Ontology

Development

• Mapping

• Sense Making

Steady State• Identify Outliers

• Model Adaptation

Copyright ©2009 BBN Technologies

Data Integration

Many of our applications focus on some aspect of data integration• Diverse databases, web services• Combining structured and unstructured

information (RDF encoding of natural language processing output)

• Translation between ontologies using SWRL plays a key role in many of our solutions

• Presentations at SemTech 2007 and 2008

8

Copyright ©2009 BBN Technologies9

AsioTM

SOAP/

RESTWS

WSDL

WSDLOntology

OWL

Mapping Ontology

OWL

SWRL Rules

RDBMS

Domain Source Ontology

OWL

QueryDecomposition

Query: SPARQL1

2

4 Data Access

3 Generation ofSub Queries

6Query Result Set

5

Data Source Ontology

OWL

Data Source Ontology

OWL

Semantic BridgeDatabase

Semantic BridgeWeb Service

BackwardsRule Chaining

Snoggle

Parliament

Semantic Query Decomposition (SQD)

Semantic BridgeSPARQL Endpoint

Automapper

Copyright ©2009 BBN Technologies

Entity Disambiguation

Co-reference resolution among multiple incomplete data sets

• Which of 500 references to John Smith refer to the same person?

Scalable use of owl:sameAs and owl:differentFrom

Automated or semi-automated

10

Copyright ©2009 BBN Technologies

Geospatial Semantics

Geospatial Semantic Web R&D program at NGA– Geospatial ontologies (including trade study)– Geospatial indexing of knowledge bases– Semantic annotation of georegistered imagery

Active participation in– Spatial Ontology Community of Practice

(SOCoP)– Terra Cognita workshop series– US Geospatial Intelligence Foundation (USGIF)

11

Copyright ©2009 BBN Technologies

Parliament

Memory-mapped embedded triple store

• Descended from DAML DB

• Used in most BBN projects

• Soon to be released as open source

http://parliament.projects.semwebcentral.org

Copyright ©2009 BBN Technologies

Dynamic Data Visualization

Challenge: Improve our ability to identify patterns, relationships, and exceptions in large, complex data sets

Solution: Capitalize on the innate ability of people to discern motion to highlight patterns, relationships, and exceptions

Results: Enables users to identify and uncover important clues and patterns buried in complex, multi-dimensional data

Applications: Intelligence, epidemiology, public safety,medical imaging, fraud detection

Complex Data Visualization and Analysis

Scatterplotvictim vs. cityScatterplot

victim vs. city

Event distribution in spaceEvent distribution in space Active Timeline HistogramActive Timeline Histogram

Highlightedevents

move in alldisplays

Highlightedevents

move in alldisplays

Scatterplotvictim vs. cityScatterplot

victim vs. city

Event distribution in spaceEvent distribution in space Active Timeline HistogramActive Timeline Histogram

Highlightedevents

move in alldisplays

Highlightedevents

move in alldisplays

Copyright ©2009 BBN Technologies

Knowledge Streams

aggregationaggregation

persistentqueries

persistentqueries

augmentationaugmentationcontextfilter

contextfilter

alertsalerts

correlationcorrelationtranslationtranslation

inferenceinference

distributiondistribution

DataDataSourcesSources

Distribution And Processing ElementsDistribution And Processing Elements

UsersUsers

CEPCEPNLPNLP

Sensor Sensor NetworkNetwork

ImageryImagery

RSSRSS

IMIM

GazetteerGazetteer

SensorSensor

Semantic Semantic WebWeb

DatabaseDatabase

Persistent pipelines• Streams of statements comprising

object subgraphs• URI naming allows drill-down

• Provenance, timestamps

Processing elements •Consume and produce subgraphs •Multiple functions may be combined

ArchiveArchive

User 2User 2

User 3User 3

Community of Interest 1

Community of Interest 2

User 1User 1

A new, highly ambitious effort by Vulcan Inc., which also develops Semantic MediaWiki+ and venture invests in sem tech. Multi-institutional.

Addresses fundamental requirements for scaling Semantic Web to widely-authored Very Large KBs in business and science that answer questions, proactively supply info, and reason powerfully.

A knowledge representation (KR) system: reasoner, language, UI, interchange.

Radically extends the KR power of W3C OWL, SPARQL, and RIF – and of SQL

Defaults and robust conflict handling – cope with knowledge quality and context

Higher-order and flexible meta-reasoning – elevate meta-data to meta-knowledge

Actions and events, cf. production rules and process models – activate knowledge

Use cases in business policies, ontology mapping, e-commerce, biomed.

Redefining the KR playing field for semantic web, business rules, and rule-based process management.

How You can be Involved:

In ~ fall 2009: try out SILK software; comment on SILK language design

Visit the SILK webpage, subscribe to (rare) announcements, cooperate in other ways

“Semantic Rules Take the Next Big Step in Power”

http://silk.projects.semwebcentral.org

Catch the SemTech-2009 talk Thurs. 6/18 9:45-10:45am

presented by Benjamin Grosof, Vulcan Inc.

Copyright ©2009 BBN Technologies

bbn.com

asio.bbn.com parliament.projects.semwebcentral.org iswc2009.semanticweb.org

For More Information


Recommended