Why ontology?

June 2017
 
Institutions need to collect, organize, preserve, and share the scholarly work produced by their people.  When people consider VIVO, they discover the VIVO ontology, and its associated ontologies.  And then people have questions.
 
Why don’t we organize information as in almost all other systems, storing data and using it to make pages or answer questions?  Why must we learn about ontologies?  Why must we map our data to ontologies? What do we get from using ontologies?
 
The first thing we get by using ontologies is meaning.  That's what ontologies do -- they provide meaning.  In VIVO, all data is represented using elements of ontologies.  These elements define the meaning of a recorded data value.  Elements provide meaning for the kinds of things in VIVO.  Other elements provide meaning for the relationships between things.  All elements are defined and the definitions are in the ontology.  The ontology is machine readable and available on-line.  Contrast this to traditional systems.  Data elements are stored in "columns."  Columns have names.  The names may convey meaning.  The meaning may be written down in documentation.  The documentation may be available.  In VIVO, the documentation is the ontology and is part of the system.  The system can not function without its definitions, that is, without its meaning.
 
We get precision.  In VIVO, we don't put meaning in text, or labels, or data structures, or software.  We put meaning in ontological elements.  We have types of things — types of organizations (university, company, and so on), types of people (faculty member, staff member, librarian) and types of many other things.  We have precision in relationships between things.  And when one VIVO site expresses its data with the ontology, other sites know precisely what is meant by the data.
 
We get consistency.  In VIVO, a date is an individual defined in the ontology.  It behaves in a way defined by the ontology.  And all things defined as dates behave the same way.  In traditional systems, columns containing dates may be processed as dates.  And in large traditional systems, it may be difficult to identify all the columns that contain dates.  In VIVO, dates are defined as dates in the ontology and can easily be found.  They must be processed as dates.  This same consistency applies to web addresses and every other kind of thing in VIVO.  VIVO treats the thing as it is defined, and can find things based on how they are defined.
 
As a result of the use of ontologies to insure meaning, and the consistency that results, we achieve commonality -- the ability to share and use each other's data.  We do not need to convert from one set of columns to another another.  We do not need to introduce additional specificity.  We can simply share, the work to share is already done.  Tools built by one VIVO site can be used by another because they share common ontologies, they share common meaning, and common representation.
 
We get multi-language capability.  Because the meaning is in the ontological elements, we can label these elements using any language we choose, and as many languages as we choose.  We do not express meaning in the labels of things, but in the ontological elements and structures of things.  The label for one’s position might say “Associate Professor” and this may be translated in many ways.  But the position itself has a type, and that type is FacultyPosition.  The FacultyPosition may have many labels, each in another language.
 
We get extensibility.  Vitro provides an ontology editor that can be used in Vitro and VIVO to add ontological elements, and thereby extend what Vitro and VIVO can represent, that is what VIVO and Vitro an “talk about.”  Additional precision can be added when that is desirable.  New types of things can be added.    New relationships between things can be added.  Entire ontologies developed by others can be added.  And unlike many systems, these changes can be made at any time, without having to bring the system down, or otherwise interrupt users of the system.  And while VIVO is focused on scholarship, the underlying general nature of Vitro supports talking about anything that has an ontology, and that could be anything.
 
We get inference.  Inference is the ability of Vitro and VIVO to  know something because we have said something precise in the context of an ontology.  For example, we might say “Mike is a faculty member”.  The ontology says “All faculty members are people”.  Therefore, Vitro and VIVO treat Mike as a person.  We don’t need to say that Mike is a person. Inference is a very powerful reasoning capability unique to ontology based systems.  In VIVO, inference is used to avoid having to enter additional data.  In the future, inference can be used to simplify queries, improve visualizations, and reduce the size of data collections.
 
And finally, we get a world community of practitioners describing scholarship and other domains.
 
We also get difficulties.  Ontologies are a new idea for most people.  There will need to be time for learning.  Training may be hard to find.  Tools are not as common as for other types of systems.  Tools are not as mature.  But support and use of ontologies is growing.  Progress is being made on all these concerns.
 
We welcome everyone to work together to refine and extend the VIVO ontologies, and to align ontology work in the VIVO community with others interested in representing scholarship.
 
Why ontology? Ontologies are the key to our ability to collect, use, and share data about scholarly activities across the world.
 
 
preserve