Linked Data: Where is the Low-hanging Fruit?
Here are my thoughts on some generic considerations, some mentioned in the recent SemTech meeting and some I jotted down following the CETIS Conference, on where low hanging fruit may be found. NB these are "generic" and not specific; the idea is that they might be useful in judging the likelihood of success of some specific good/cool/potential ideas. I am referring here to exposure of Linked Data on the public web. In no particular order:
- Ariadne's Thread. Does the current state of (poor) information management present a problem and is there resolve to find your way out of the maze? If it has become necessary to sit down and get your domain model straight and re-organise/re-engineer (some of) your information systems then you have done most of the hard work necessary for exposing Linked Data (i.e. Open to some degree) and you could usefully adopt Linked Data principles for private use.
- Ox Pecker. Is there a mutual benefit between you and another data provider? Can this be amplified by explicit technical, financial or effort (etc) support one or both ways? This builds on the essential attribute of linking.
- Sloping Shoulders. Can you avoid creating an ontology? No-one else will care about it if you do.
- Aspirin. Does anyone have a headache that can be made better? Is there an institutional/business problem that can be solved? (this is not the same as Ariadne's Thread)
- Blue Peter. Is the creation or acquisition and processing and dissemination of information already something you do? Is the quality and availability of the information something you invest effort in? This is a ready-made candidate for Linked Data.
- Cow Path. Is information you already make available (as web pages or PDF etc) used by others in ways you know about and understand?
- UFO. Do people want to refer to something you have or do but don't have an unambiguous way of identifying what they are referring to? Could you provide a URI for the thing and information about it?
- 2+2=5. Is there clear value to be gained from linking the information that is to be exposed? Can people do something new, do they want to and will they continue to want to?
- Chatham House. Avoid exposing data that identifies, or could identify, a person.