RDF is an open WWW standard originating in 1999, devised to remove the boundary between knowledge and data. By adopting RDF you can process knowledge just as you process data, and you can deploy your data as knowledge.
RDF is based on the idea that all knowledge and data can be decomposed into individual facts, like matter into atoms.
The facts are called triples because they have three parts:
| 1: Subject | 2: Predicate | 3: Object |
|---|---|---|
| Topic of interest |
A relationship | Related topic |
| Tom | Likes | Football |
Some triples link subjects to simple characteristics rather than another topic:
| Subject | Predicate | Object |
|---|---|---|
| Tom | was born | 1980-10-01 |
If you use consistent names for your topics ('resources' in RDF) and your predicates or relationships then you can model all your data/knowledge in one big list of facts, also known as a semantic graph, and store them in a triplestore.
You can then find all you need to know about Tom in all the facts with subject or object Tom.
This is a very simple idea with many implications.
The developers of RDF have addressed those implications over time to create a range of open standards which are widely implemented.
RDF is now an important presence under the surface of the world wide web. It's free for all to use.
You can hear what the BBC, a major user, have to say about it.
RDF for surveys
Using RDF you can store everything about your survey process as triples, including:
- Data about surveys: e.g. sponsor, sample structure, title
- Questionnaires: questions, answer lists, language alternatives
- Responses: every response to every question
- Results: every cell of every cross-tabulation
- Subject specific data e.g. brands, product fields, media
The same triple store can hold all your projects so that you can query any of this data across all surveys, track the participation of respondents across surveys, the history of questions in longitudinal studies, and much more.
The intelligence moves from the software to the data and all software operates from the same data. The need for major specialised suites for software for different phases of your work or for different types of project can be greatly mitigated, and data can be reused over time or between applications much more easily.
X-MRdf have a demonstrator project for RDF in survey research. Please contact us for more information.