At ReadWriteWeb, Richard MacManus notes that neither online “open data” on its own, or online “linked data” on its own, is terribly useful. The real holy grail is online data that is open and linked. When this becomes pervasive, knowledge seekers will be better able to quickly survey that slice of the landscape they want to view, and then drill down, aggregate or add on where they wish. This without the piecemeal hunting for reliable data groupings that’s caused partly by unsophisticated or non-existent reliability indicies, and also current Web search protocols which tend to be keyed to exact terms rather than intended meanings. Linkage of open data will occur through use of “meta-deta” or descriptive terms of data categories.

This will allow search engines to capture broader meanings that link data and data sets, and is one core component of what is being called “The Semantic Web,” or “Web 3.0.” But it will require far-reaching agreements among data-providers on “master category” descriptors and their thorough application to online data. It’s not going to happen overnight, but business, government and non-profits will all find it in their best interests to collaborate on strengthening the ties that should bind data and allow filtering by reliability.

The initial “relational” piece is fairly huge in terms of knowledge organization and utilization. In fact, one of the real inventors of the Internet, is making this a priority.

The British government has invested £30 million (US$45 million), in a research center to further develop Tim Berners-Lee’s Semantic Web. The center, to be called the Institute for Web Science, will be run by Berners-Lee, who formulated the basic protocols for the Web, along with University of Southampton artificial intelligence professor Nigel Shadbolt…The Semantic Web is Berners-Lee’s vision for how the Web should evolve beyond its origins as a worldwide repository of human-readable hyperlinked documents…

By using linked data, machines should be able to make inferences and reason about data found they find the Web, without human intervention, in effect turning the Web into a worldwide database. Linked data relies on a number of still-emerging Web standards. One is RDF (the Resource Description Framework), which can link two disparate sources of data…Linked Data, a compendium of linked data sources, has counted over 13 million triples, or RDFs that connect two different sources of data, from 200 data sources.

Why is this so huge, potentially? Suppose there was a global advocacy organization seeking to raise large sums of money for establishment of secular K-12 schools and universities in the developing world and particularly Islamic-majority nations, emphasizing liberal arts, free inquiry and science-based methods. And suppose that to make a pitch to very well-heeled and willing funders, the organization needed in its prospectus to summarize the most current data on several dozen nations, for estimated percentage of population within a given age range who were even enrolled in school, plus how many students were in primarily religious versus secular schools.

Other researchers have covered the topic in bits and pieces, and some have attempted overviews in recent years. None of this is deemed good enough, however, and Web searches return largely unfiltered results. With the semantic Web done right (and this includes user profiles and data-ranking preferences baked into search parameters) results could be quickly winnowed to data sets provided by designated A-list sources sources such as top NGOs, universities, and primary global policy and research institutions like the United Nations and the OECD. This in turn would lend greater authority to data retrieved, and more quickly reveal next steps needed to complete the proposal, make it convincing, and win support from donors.

It’s not that much of the knowledge isn’t already out there. It is. The issues are organization, access, reliability and utility.