Personalization in the EPOS project - SRI Internationalnysmith/slides/aic-seminars/060907-sauer... · 2006-09-25 · Manager Project Contract Offer Company basic superclasses Rep

Semantic Desktop Overview

SRI SeminarThe Nepomuk Project &

Social Semantic Desktop & current state of the gnowsis

http://[email protected]

Leo [email protected] created using http://www.sp-studio.de/

http://www.sp-studio.de/






Who am I?


Leo Sauermann

• Researcher at the Knowledge Management Lab of the DFKI German Research Center for Artificial Intelligence

• Austrian, born & studied in Vienna• Created the Gnowsis Semantic Desktop (Diploma

Thesis) - 2003• Preparation for and now working on the Nepomuk

project - 2006• I need the Semantic Desktop to write down notes

and its a step to cyberspace++• Jack Park and I are in the organization committee of

the Semantic Desktop Workshop


This talk

• Writing down information• PIMO – Ontology system• Nepomuk Project• Gnowsis Project• Walkthrough of the system (and wiki)• Input for discussion

Nepomuk VS OpenIris/Calo


question:


How can I write down information?


Information: NepomukNepomuk is an EU project . It’s topic is the Social Semantic Desktop .DFKI coordinates .


To write down, I need language


NepomukNepomuk is an EU project . It’s topic is the Social Semantic Desktop .DFKI coordinates .

A language of the Semantic Desktop

• Terms• Types• Relation types• Relations

• RDF


In RDF@prefix leo: <gnowsis:[email protected]/pimo>.

@prefix pimo: <dfki.de/.../pimo>.

@prefix project: <example.org/projects>.

leo:Nepomuk a project:EUProject; project:coordinatedBy leo:DFKI; pimo:hasTopic leo:SemanticDesktop.

NepomukNepomuk is an EU project . It’s topic is the Social Semantic Desktop .DFKI coordinates .


PIMO is The “Personal Information Model”

• To write down information, we use RDF for the data and and RDF/S ontologies for the semantics

• A stack of vocabularies and ontologies• used by the – user

– creates instances– creates classes & properties (on the fly)– annotates resources– downloads new vocabularies and ontologies– exchanges data with others


domain-independentSemDesk Upper Level

Person Role Document OrganizationTime

domain-independent, adapted to Semantic Desktop and Nepomuk

SemDesk Mid-Level

Manager Project

Contract

CompanyOffer

basic superclasses

Rep Lang

SystemItemsThing

onto

logy

impo

rtsMessage

dfki.de/ont/pim/pimo

PIMO vocabularies / ontologies• PIMO-Basic

defines the basic language constructs.

• PIMO-UpperA domain-independent ontology defining abstract sub-classes of Thing.

• PIMO-Mid: More concrete sub-classes of upper-classes. The mid-level ontology serves to integrate various domain ontologies and provides classes for Person, Project, Company, etc.


Organizational Structure

Domain Model: Bibtech A

Heiko Car-Ent Report56

Report

EPOS

PIMO domain ontologies• Domain ontologies

A set of domain ontologies where each describes a concrete domain of interest of the user.

• Vocabularies– GEO– Beer Ontology– DOAP

• Shared Ontologies– The user’s company and its

organizational structure– a shared project– data collection (SUMO,

gene ontology)


basic superclasses

Rep Lang

SystemItems

domain-independent

SemDesk Upper Level

Person Role Document OrganizationTime

domain-independent, adapted to Semantic Desktop and Nepomuk

SemDesk Mid-Level

Manager Project

Contract

CompanyOffer

representing extracted data in RDF/Smultiple vocabularies

Native Data Vocabularies

vCard

vEvent dublin core

foafimagePerson Image

Thing

sub-classes

Organizational Structure

Domain Model: Bibtech A

Heiko Car-Ent Report56

Report

EPOS

onto

logy

impo

rtsMessage

aperture.semanticdesktop.org/data

dfki.de/ont/pim/pimo

all PIMO ontology layers


Paul

Paul‘s PIMO - Personal Information Model

personal information model of one user Imports all other ontologies and defines extensions

Ontology side Native Resources

PIMO of Person:Paul

Paul

Project ZReport41

File X e-mail2

vCard H

Rep Lang

SemDesk Upper Level

SemDesk Mid-Level

Native Data Vocabularies

Domain: Bibtech ADomain: Paul’s company

Paul imports all of them


State of PIMO in Nepomuk

• Nepomuk currently works on the Nepomuk Representational Language (NRL). – extensions to RDF/S. Pruned - limited features– inverse prop, Metadata– will be published soon (2006)

• upper, mid, and domain ontologies are expressed in NRL– will be published later (Dec 2006?)


Ontologies with quality

• peer review• formal description• good URLs• written documentation• examples• validation rules


Integration: The Social Semantic Desktop


From existing structures to PIMO

• Existing data is expressed in RDF– documents, files, e-mails,

address book, ...– represented as RDF

• Matching & Creation Algorithm“Rebirth Machine”

• Personal Information Model– Personal Concepts

• Topics• Places• People• Types• Workflow

• with relations to files and folders


Nepomuk Desktop Data Services

• Two example implementations:– Aperture– Beagle++

• a service to extract data, a middleware for data integration on the desktop

• used to crawl data for desktop search

• description of this standard is due Dec 2006


Nepomuk


Goal: Build the Social Semantic Desktop

Desktop: Help individuals in managing information on their PCSemantic: Make content available to automated processing Social: Enable exchange across individual boundaries

colleague

friend

acquaintance

NEPOMUK enabledpeers

Personal Semantic Web:a semantically enlarged intimate supplement to memory

Social protocolsand distributed search

EmailPerson

Topic

WebSite Document

Image

Event

Person


NEPOMUK will deliver …

• The freely available open-source framework comprises– Standardized interfaces, architecture, ontologies– Plug-in architecture, extendable to suit application needs– Building upon standard semantic web technology

• A set of reference implementations covers representative application domains– Consulting, BioScience, Strategic planning, Distributed Community work

• A vivid community ensures uptake and extension beyond the project‘s duration– Active scientific community (http://semanticdesktop.org)– Involvement of open source developer communities– Setup of user groups and domain-specific application communities

http://semanticdesktop.org/


•Internal exploitation •Initial awareness via open source and scientific communities

Exploitexternally

After 3-5 years3

NEPOMUK

1After first 18 months

(initial project results are available)

2 After 2-3 years

Exploitinternally

•External application of project results by industrial partners

•Uptake in the scientific community

•Wide-scale awareness within open source community

•Commercial exploitation by dedicated spin-offs

•Specific product development

•Wide societal impact

Individual exploitation concentrates on uptake, adaptation, and commercialization by tool adaptors & consultation services, i.e. spin-offs

Collective exploitation will focus on establishing & maintenance of a vivid scientific society

Pursuing standardization and platform & prototype development Collective support of standardization efforts (e.g. W3C working group)

NEPOMUK realizes the basis for manifold exploitation


The NEPOMUK consortium unites industry, research, and representative users


NEPOMUK is a three-year Integrated Project within the EU 6th Framework Programme

• Project FP6-027705 NEPOMUK• Duration: 1.1.2006 – 31.12.2008• EC contribution: 11.5 M€ (Total budget: 17.1 M€)

• For further information, please contact– Project website: http://nepomuk.semanticdesktop.org– E-mail: [email protected]– Coordinator: Dr. Ansgar Bernardi

Erwin-Schrödinger-Str.67663 KaiserslauternGermanyTel. +49-631-205-3583

http://nepomuk.semanticdesktop.org/





mailto:[email protected]





gnowsis prototype


Gnowsis is a prototype of predecessor projects

• service oriented architecture

• integrates various applications• semantic wiki• continous development, Java, Sesame2,

Jena, ontologies, PIMO• beta 0.9.1 version

• walkthrough


Semantic Applications

Desktop Search

GnowsisServer Aperture

CrawlersOutlook

e-mail server

filesystem

Outlook

PIMO Editor

Crawler

Ont. Matching

Files

e-mail

Gui invocationTagging

Clustering

Desktop ApplicationsApplication Plugins

Sesame2 Repository

Resource Store

PIMO Store

Configuration Store

Service Store

Domain Ontologies

Lucene Index

Personal Wiki

Web 2.0 Interfaces


PIMO is filled

• automatically from data

DataPaul‘s files & e-mails

RDF Database

adapter to RDF

matchingto

ontology

PIMO Store

aperture.sf.net(check it out!)


Aperture.


Aperture• A java framework for getting data and

metadata• extracts full-text content• and metadata• from various file formats• and structured data sources

– IMAP– Outlook– flickr.com



related work



Features of IRIS


Iris Layers


Clusteringemail is colored - each word colored to indicate cluster.

general sense of how this email relates to one or more projects.


Similarities to Nepomuk

• cognitive Assistant (CALO) - similar goal• all AI that is possible:

– LSA, DSA, clustering, NLP …• focus on PIM• annotations, RDF, Ontologies

IRIS: Integrate. Relate. Infer. Share.Adam Cheyer and Jack Park and Richard Giuli

Proceedings of the 1st Workshop on The Semantic Desktop 2005


Differences to Nepomuk?

Nepomuk has• service oriented architecture• peer to peer and social exchange• consortium, review and standardization

processOpenIRIS has• tested and useful AI scenarios• a working prototype• a core team on-site (=effective)???


Science Fiction

UI groupfancy Semantic UI

p2pcollaboration

KM groupgreat KM system

EAI groupadapters

Web Services groupgreat protocol


Summary

• writing down information• PIMO – Ontology system

• nepomuk Project• gnowsis Project• walkthrough of gnowsisLets discuss:• Nepomuk VS OpenIris/Calo


Thanks, Questions?

You listening

Jack Park organizing today’s seminarNepomuk people sharing the ideaGunnar Grimnes sharing the workAnsgar Bernardi managing me

BMBF funding parts of thisIST funding most of thisIngrid Brunner (wife) Loving MeJesus Loving Me+


Application: Drop Box• Helps filing information• uses PIMO structures

concepts and folders• uses DOCS for text similarity• Knows the users model and is

trained by using it• process flow

– files are stored into a Drop-Box folder– files are text-analysed and possible

target folders are suggested– Drop-Box user interface shows– user selects a folder, classify– files are moved and classified

• not used, but obvious:OBS – current context

doc:… EPOS …

Project:EposEPOS, DFKI,Maus, …

PIMO of Paul

=


Weird Experiments with gnowsis: alternative user interfaces

Foafnaut

Ultima Online

Fenfire


Context Assistance

• Sidebar• can be switched off• shows current context

– documents– people– projects– topics

• changes dynamically• use: open related information, pro-active,

non-obtrusive assistance system


Application: Drop Box• Helps filing information• uses PIMO structures

concepts and folders• uses DOCS for text similarity• Knows the users model and is

trained by using it• process flow

– files are stored into a Drop-Box folder– files are text-analysed and possible

target folders are suggested– Drop-Box user interface shows– user selects a folder, classify– files are moved and classified

• not used, but obvious:OBS – current context

doc:… EPOS …

Project:EposEPOS, DFKI,Maus, …

PIMO of Paul

=


Semantic Search• search over EPOS data (PIMO)• can be personalized using rules SPARQL queries• example

# found something? -> infer other representations via SPARQL(?hit retrieve:item ?x) -> querySparql('CONSTRUCT { ?x

pimbasic:hasOtherRepresentation ?y } ')

# found a project? -> also show members(?hit retrieve:item ?project),(?project rdf:type org:Project) ->querySparql('CONSTRUCT { ?project org:containsMember ?m.}).

• Innovation– search result expansion using SPARQL– customized rules for search – only when word “x” is searched, include

these results, etc


search


Paul

Context Service

• Plugins gather user actions• Elicitation of task concepts• Notification of GUI

Bayesian Network

UA UA

Domain

Wf Task

NOP NOP NOP NOP

Wf Task

TaCo

Domain

UA

NOP

PIMMaus

PIM MidDFKI KM

PIM Upper

PIM Basic


The nepomuk project• Different research institutes are working on a vision titled "Semantic

Desktop", a semantically enhanced desktop computer that allows us to access semantic web data and desktop data in a uniform way. The European Union Integrated Project NEPOMUK (http://nepomuk.semanticdesktop.org) started in 2006 and intends to realize and deploy a comprehensive solution – methods, data structures, and a set of tools – for extending the personal computer into a collaborative environment, which improves the state of art in online collaboration and personal data management and augments the intellect of people by providing and organizing information created by single or group efforts. NEPOMUK brings together researchers, industrial software developers, and representative industrial users. In this talk you will get an introduction on the theory behind the Semantic Desktop, ontologies, databases, user interfaces and projects that work on this topic. Details about the current open-source implementations are presented and a demo is given. The lecture will finish with a discussion, where similarities and differences to the OpenIRIS project by SRI will be an important question.



Documents

Personalization in the EPOS project - SRI Internationalnysmith/slides/aic-seminars/060907-sauer... · 2006-09-25 · Manager Project Contract Offer Company basic superclasses Rep