In several essays from 2009, I tried to explain ways that news organizations could profit from building information models that would store information in potentially valuable datasets. These essays include "The Lack of Vision Thing," "The Imagination Gap," "The Future is Closer Than You Think," "Narrative is Dead! Long Live Narrative," "Free Wants to be Big," and "The Limits of Social." I examined these concepts in a larger context in December 2010 in "Imagining the Semantic Economy."
This idea recognizes that news stories represent distinct information models, as in an obituary contains required information that is different from a wedding announcement or a real estate transaction. The required elements of these models can be expressed a structured and semi-structured data, and if captured in a machine-friendly way, the resulting dataset can become a vaulable asset.
In some instances, the value you will in terms of unique content that only the data-gathering organization can publish profitably. But given the right tools and quality control standards, a dataset of information about residential house fires in a community can become a dataset that could be marketed to insurance companies, home-improvement retailers, personal security companies, etc. This persistent value in information would provide an alternative to our current system, in which interestingness -- an ephemeral quality at best -- is the only dollar value assigned to news.