1
2
3
4
5
6	"Andy Goldsworthy" Andy Goldsworthy
7
8
9
10
11	Web site targeted to NLSI needs: VIVO Collaborative, scholarly initiative of librarians across Cornell Provides well annotated, organized info on: bibliographic resources biology/bioinformatics databases Cornell life sciences: people, research, courses, services/facilities, events/activities Encourages faculty/research staff collaboration across Cornell
12	Typical Search Engine
13	VIVO
14	How is VIVO Different? No fixed collection Does not mirror organizational structure Provides context for information Fluid: - will go in directions we cannot now anticipate - adapts to content domains beyond our control
15	What Does this Gain Us? Simplicity – everything displays consistently Context – categorized browsing and searching Fluidity - users can roam, not just drill down Flexibility to grow and change in response to content and use
16	VIVO: http://vivo.library.cornell.edu
17	Nuts and Bolts Design Requirements Sustainability Ontology underpinnings Strategies for moving forward General applicability
18	Design Requirements for VIVO Lightweight Flexible Extensible Sustainable Compliant with standards Amenable to automation
19	Sustainability Accessibility – immediate curator feedback Granularity – not too detailed Quality – commitment to get it right Flexibility – ability to adapt Proactive alignment with emerging standards/registries/shared resources in the library and life sciences communities
20	Phase One Approach to Updating Downloading recent publications Cultivating departmental contacts Emailing “review your entry” solicitations to individuals Dead link checking
21	Vivo as Harvester
22	Vivo Antecedents FRBR
23	Extending the Scope FRBR
24	ABC Class Hierarchy with Property Relationships
25	Innovations Database structure Context-specific editing Curator participation in content organization and presentation Support for migrating content Ability to incorporate controlled vocabularies and domain ontologies
26	Typical Database Organization
27	Consolidation
28	Ontology Framework
29
30	Properties are Relationships
31	Vivo Entity and its Property Relationships
32	Relationship Glut How do you avoid an unmanageable tangle of property relationships among so many classes?
33	Connect Only What You Need To Following the W3C OWL model: Properties are defined between specific classes Properties are inherited by subclasses but not necessarily switched on
34	Adaptive Structure Data are added to the site in response to interest and leads Curators add more types and relationships to support differentiation where entries accumulate The scaffolding adapts to areas of greater structural load
35	VIVO as a Harvestable Resource Becoming an OAI-PMH service provider Selection by type of entity Selection by property/relationship based on DC, MODS, or any preferred standard VIVO web services Search response with name, type, and URL Further options Additional context Inferencing results Provenance or annotations
36	Extending Vivo Filtering By major subject area, college, campus Improved searching Augmenting terms via a choice of external thesauri Distributed editing Departmental liaisons Templates to populate relationships Export RDF, METS Preservation, pruning, and migration Visualization
37	Generalizing VIVO Modular design Clone with or without ontology Harvester and harvested Expose metadata from multiple digital library collections in a coherent, accessible framework Collaborations With other aggregations of metadata With domain-specific resources
38