jump-start your taxonomy project - synaptica llc · 2018-01-03 · linked open data taxonomy...
TRANSCRIPT
©2017Synaptica,LLC www.synaptica.com
Jump-start Your Taxonomy Project
Dave ClarkeCEO Synaptica
@synaptica
TaxonomyBootcampLondon2017 #tbcl17
©2017Synaptica,LLC www.synaptica.com
Jump-start Topics
3 things to help jump-start a taxonomy project
1. Know How• Management & stakeholder buy-in• Industry standards• Take an audit of your existing content & systems• Whiteboard your knowledge domain• Adopt & adapt ontologies• Design your taxonomy scheme(s)
2. Starter Taxonomies• Leveraging Linked Open Data resources• Licensing third-party taxonomies
3. Taxonomy Management Tools• Use a taxonomy management tool• Share work-in-progress with stakeholders often25MinuteSession
©2017Synaptica,LLC www.synaptica.com
Number 1 - KNOW HOW• Management & stakeholder buy-in• Industry standards• Audit existing content & systems• Whiteboard your knowledge domain• Adopt & adapt ontologies• Design your taxonomy scheme(s)
©2017Synaptica,LLC www.synaptica.com
Ahren LehnertSenior Manager Text Analytics, Synaptica
“Whatistheplotofmystory?” Youneedagoodstorytoselltaxonomytouppermanagement,communicatethevalueandtheeverydayusetoyourendusers,andtopresentyourprojectinanice,compact,elevatorpitchcleartoalllevelsandrolesin
theorganization.”
REACH OUT TO THE LIBRARY AND INFO PRO MARKET
9 | COMBINED MEDIA KIT 2017 INFORMATION TODAY • COMPUTERS IN LIBRARIES • ONLINE SEARCHER
ONLINE SEARCHER EDITORIAL CALENDAR 2017Information Discovery, Technology, Strategies
Jan/Feb › International Policies › Digital Humanities › Talking Among Ourselves › Bonus conference distribution: American Library Association Midwinter (ALA)
Mar/Apr › Open Source Solutions › Scholarly Digital Social Sciences › Financing the Stone › Bonus conference distribution: Computers in Libraries ’17 (CIL),
Association of College & Research Libraries (ACRL)
May/Jun › Big Data and Library Research › Public Health › Toy Story › Bonus conference distribution: Data Summit, Special Libraries
Association (SLA), American Library Association Annual (ALA), Medical Library Association (MLA), Association of Independent Information Professionals (AIIP)
Jul/Aug › Getting GLAMerous › Legal Limits › Text and Data Mining › Bonus conference distribution: American Association of Law Libraries
(AALL)
Sep/Oct › Artistically Searching › Taking an Analytical Approach › Dealing Libraries a Data Hand › Bonus conference distribution: WebSearch University, Internet
Librarian International, Internet Librarian, and Charleston Conference on Collection and Development
Nov/Dec › Search Goes Back to the Future › Taming Taxonomy Terrors › Global Perspectives › Bonus Conference Distribution: KMWorld, Enterprise Search and Discovery
Editorial content focuses on subjects of topical interest to librarians and other information professionals:› Science & Technology› Business & Finance› Medical & Pharmaceutical› Social Sciences & Humanities› News & Current Awareness› Legal, Tax, Regulatory, & Intellectual Property› Competitive Intelligence › Usability
In every issue:› Information Literacy› Internet Technologies› Open Access, Open Source› Business Research› Conference Coverage› Information Industry News› Library Website Design› Book Reviews
Contact David Panara • Advertising Sales Director • (609) 654-6266 ext. 146 • (609) 257-0112 • fax [email protected]
Management Buy-In
©2017Synaptica,LLC www.synaptica.com
• Segmentallpotentialusergroupsandtalktothem
• Describetheirneedsandhowtheywillusethetaxonomy
• Documentthesystemstheywillbeusingandtheuserexperiencetheywillneed
Stakeholder Buy-In
©2017Synaptica,LLC www.synaptica.com
https://www.iso.org/standard/53657.html www.niso.org/standards/resources/Z39-19.html
TheBible AnOlderTestament
Lexicographicstandards Semanticwebdatastandards
https://www.w3.org/TR/2009/REC-skos-reference-20090818/
SKOS
Industry Standards
©2017Synaptica,LLC www.synaptica.com
KMKnowledgeAudits&Mapping
What comes before taxonomy – knowledge audits
©2017Synaptica,LLC www.synaptica.com
[email protected] ourdiscoveryquestionnaire
WhyAuditExistingSystems?Because…
“Whenyouautomaticallyormanuallyselectmetadatatoclassify yourcontent,itcomesfromthetaxonomy.Whenyoutypeintoyoursearch boxandtherearetype-ahead suggestions,they comefromyourtaxonomy.Whenyoulookattheglobalnavigation,thevaluescomefromthetaxonomy.Whenyouapplytextanalyticstocontenttodeterminewhatitisaboutortoclusteritemsbytopic,classificationtermscomefromthetaxonomyandextractedconceptsgobackintothetaxonomy.”
Ahren Lehnert,fromtamingTaxonomyTerrorsinInfoToday’sOnlineSearcher
Systems audits
©2017Synaptica,LLC www.synaptica.com
UserGroups NeedsTaxonomycurators Import &adopt
Build&maintainHumanindexingguidelinesMachineindexingrulesWorkflow&governancePublicationcycles&versioncontrol
Contentcurators SME&stakeholderreviewContentnavigation
Human&machine-aidedindexers
Search&browseAccessguidelines& rulesSubmitcandidates
Search&discovery ResolveambiguousconceptsMap userquerylanguagetoindexinglanguage&Submitcandidates
qHumantaggingbycontentauthors
qHumanindexingbyprofessionalindexers
qMachine-aidedindexingsystems
qCatonomydevelopersqFacetedend-usernavigationinterface
qSearch&query-refinement
What comes after taxonomy – indexing and categorization
©2017Synaptica,LLC www.synaptica.com
Withdrawn
Published
Quality Control
Work in Progress
Candidate Unapproved Deleted
Approved Deleted
Published
Withdrawn Deactivated
“Taxonomygovernancecomprisesthepolicies,procedures,anddocumentationfortheongoingmanagementanduseoftaxonomy.…theexistenceoftaxonomyisitselfaformofgovernance.”
HeatherHedden– AccidentalTaxonomisthttp://accidental-taxonomist.blogspot.co.uk/2013/12/taxonomy-governance.html
Workflow and governance
©2017Synaptica,LLC www.synaptica.com
2. Describethebusinessprocessesandfunctions thataninformationsystemneedstosupport
3. Choosetheformaldatamodelssuitedtotheinformation,processes,databasesandITsystemstobesupported
Getting Started with Domain Models and Process Models
1. Describethetypes ofreal-worldobjects andabstractideas thataninformationsystemneedstoreference
People
SubjectsPlaces
Literature
ISO25964
©2017Synaptica,LLC www.synaptica.com
Diagram the Entities, Attributes and Relationships
Descriptiveinformationaboutreal-worldobjectsandabstractideascomprisesthreegenericcomponents:
Entities(ovals)
Uniquethingsandideas
Attributes(boxes)
Descriptivepropertiesofentities
Relationships(arrows)
Connectionsbetweenentities
The Tragedy of Macbeth
(LiteraryWorksScheme)
William Shakespeare(PersonNames
Scheme)
Stratford upon Avon
(GeospatialNamesScheme)
Ambition (Overreaching)
(SubjectThesaurus)
has Birth Place
April23rd1564
Birth Date
Creator of
Ambitionandtheconsequencesthatfollowwhenambitionoverstepsmoralboundaries.
Scope Note
IsAbout
52.1900°
Latitude
1.7100°
Longitude
1599- 1606
Written
©2017Synaptica,LLC www.synaptica.com
Multiple KOS Schemes
Concept Schemes collections of like things(i.e. with common set of attributes)
Entities [ovals]uniquely named things and concepts (non-preferred terms redirect to preferred terms)
Attributes [rectangles]properties of things with literal values
Relationships [arrows]named directed relationships between entities
Subjects Scheme
Literary Works Scheme
Person Names Scheme
Place Names Scheme
separate concept schemes are used to represent collections of like-things
in addition to intra-scheme relationships a multi-scheme KOS may supportinter-scheme relationships that make factual assertions about the properties of
an entity in relation to entities in other schemes – if formally defined this becomes an ontology
The Tragedy of Macbeth
William Shakespeare
Stratford upon Avon
Ambition (Overreaching)
April23rd1564
BirthDate
Author Of
Ambitionandtheconsequencesthatfollowwhenambitionoverstepsmoralboundaries.
ScopeNoteIs About
52.1900°Latitude
1.7100°Longitude
1599- 1606
Written
TheScottishPlay
Use
Birth Place
©2017Synaptica,LLC www.synaptica.com
Worked Example – From Data Model to KOS in 30 Minutes
• ConfigurefourKOSschemesforstoringthefourfundamentallydifferenttypesofthing(People,Places,WorksandSubjects)
• Foreachschemeconfigurepropertyfieldstostoretheattributesabouteachthing(BirthDate,Latitude,Longitude,etc.)
• Configuresemanticrelationshiptypes tosupporttherelationshipsbetweenthings(CreatorOf,hasBirthPlace,etc.)
Timetoconfigureentiredatamodelwithmultipleschemes,propertiesandattributes- lessthan30minutes…
©2017Synaptica,LLC www.synaptica.com
Number 2 - STARTER TAXONOMIES• Leveraging Linked Open Data resources• Licensing third-party taxonomies
©2017Synaptica,LLC www.synaptica.com
Build Buy
Newmantrafortaxonomyprojects:
ADOPT first
ADAPT second
CREATE third
©2017Synaptica,LLC www.synaptica.com
When do third-party taxonomies work well
üû
Corporate&EnterpriseTaxonomies
STEMs:Science,Technology,Engineering&Mathematics
HCLS:HealthCare&LifeSciencesCulturalHeritageNewsMediaGeospatial
PersonNames
Products&ServicesCommodities
FinanceLegal&Regulatory
©2017Synaptica,LLC www.synaptica.com
Linked Open Data Taxonomy Sources
Jump-startyourtaxonomyproject
Trustedauthorities
Manydifferentsubjectdomains
Millionsofconcepts
Manysourcesinthepublicdomain
Standardelectronicformat
Livequeryand/ordownload
©2017Synaptica,LLC www.synaptica.com
Online Directories of Taxonomies
http://www.taxonomywarehouse.com http://www.bartoc.org
©2017Synaptica,LLC www.synaptica.com
http://www.taxonomywarehouse.com/details.aspx?vunid=89115
• 115,000concepts• Professionalstandards-compliantthesaurus
• Arts,Sciences&Businessdomains• Availableasmasterthesaurus• Orin69topicalsubsets• [email protected]
Commercial Example 1: Gale-Cengage Taxonomies
©2017Synaptica,LLC www.synaptica.com
https://www.ap.org/en-us/services/planning-and-media-tools/metadata-services
• 200,000concepts• Professionalstandards-compliantthesaurus
• News-relatedtopics,PeopleandOrganizationandGeospatialNames
• DemosavailablefromSynaptica• LicensingavailablefromAPatthelinkbelow
Commercial Example 2: Associated Press Taxonomies
©2017Synaptica,LLC www.synaptica.com
Number 3 - TOOLS• Use a taxonomy management tool• Share work in progress with stakeholders often
©2017Synaptica,LLC www.synaptica.com
✔ Theyenforcelexicographicstandardspreventingmanycommoneditorialerrorsandensuringsemanticintegrity
✔ Theysupportdataexchangestandardsthatfacilitatedataexchangeandsystemsinteroperability
✔ Theyallowtaxonomyteamstocollaboratetogetherthroughrole-basedpermissions,governanceandworkflow
✔ Theyenabletaxonomiststosharework-in-progresswithotherstakeholders securingbuy-inthroughoutthedevelopmentprocess
Why use a taxonomy management tool
©2017Synaptica,LLC www.synaptica.com
Jim SweeneySenior Manager Taxonomy & Ontology, Synaptica
“Graphitepresentsanallnewgraphicaluserexperiencethatmakestaxonomyeditingchild's-play,whilealsopackinginpowerfulontologymanagementtoolsthatcansupportthemostsophisticatedknowledge
organisationsystems.”
Example tool: Graphite – you saw it first at TBCL17
©2017Synaptica,LLC www.synaptica.com
Drag-and-dropwithinthetreestructuretocreateandadjusthierarchicalstructures
©2017Synaptica,LLC www.synaptica.com
Drag-and-dropintotheSKOSNarrowerpaneltocreateanewhierarchicalrelationship
©2017Synaptica,LLC www.synaptica.com
Drag-and-dropintotheSKOSRelated paneltocreateanewassociativerelationship
©2017Synaptica,LLC www.synaptica.com
Beyond SimpleKOS – working with Ontologies and classes
©2017Synaptica,LLC www.synaptica.com
Involving Stakeholders with Granualar Project Views
©2017Synaptica,LLC www.synaptica.com
OntologyClasses supportKnowledgeOrganizationsSystemsthatcontainresourcetypeswithdifferentProperties suchastaxonomyConcepts andpublishedResources
Beyond SimpleKOS – building Ontologies and classes
©2017Synaptica,LLC www.synaptica.com
Allprizesarecompletelyfree-of-chargecloud-hostedGraphitesystemswithfulltechsupport:
1st Prize:Oneyearsystem2nd Prize:Sixmonthsystem3rd Prize:ThreemonthsystemCompetitionstartsnowandclosesattheendofTBCL- 17:00onOctober18th.
#tbcl17 #SynapticaGraphite Competition
PrizeswillbeawardedbytheSynapticateamandannouncedonThursdayOctober19th
theywillbebasedonourjudgmentofmostprofound,enthusiasticorwittyanswers.
Toenterjustreplytoorre-tweetthispostandtelluswhyyoulovetaxonomy