The Launguage of Valuation
To talk about observing the democratic interactions of the community and interpreting them to make decisions about the value of content and people in the community is to talk about solving a very hard problem. Surely, in order to make the output of something like this useful, we’re going to need a simple but powerful language with which to describe what we’re doing.
When we developed our engine, we had this reality in mind. We call the work of the engine — as it observes and interprets — “valuation“. We are “valuing” members of the community and their contributions to it. To describe this, we have developed a “valuation language” which calls on our knowledge of the English (or any other) language to help us describe what’s going on inside the CVE.
That may sound complicated, but it really isn’t. It boils down to defining every action that can be observed in the community as a sentence with the grammer, “Subject verb object” — just like the English language. Here are just a few examples…
- Visitor views content item
- Member refers member
- Member rates content item
- Member contributes forum post
In each case, as with our everyday speech, a subject (may be a content item or a member of the community, or even other entities) is performing an action on an object (content item, member, etc). As each of these sentences is defined, observed and interpreted by the valuation engine, value judgments can be made about the “nouns” in the sentence.
If a visitor views a content item, then he or she has made an implicit value judgment about that content item. The engine can observe and interpret that — capturing the information in the item’s score — and remember it to set context for future observations / decisions. If a member rates a content item, this says something about the member performing the action, as much as it says something about the content item being acted upon.
This is the heart of how Intelligent Scoring works, and a very simple but power language of valuation makes it possible.
4 Comments »
Leave a comment
-
Recent Comments
Fascinater’s W… on Terms and Definitions Jeff Block on The New Form of Media Mold Remover on The New Form of Media indianglory on The Launguage of Valuatio… Jeff Block on The New Form of Media
-
Archives
- August 2006 (4)
- July 2006 (10)
-
Categories
-
RSS
Entries RSS
Comments RSS
Jeff,
Your ‘Subject verb object’ definition is very similar to RDF’s idea of a “triple” ( http://www.xml.com/pub/a/2001/01/24/rdf.html?page=2 )
Was this intentional and you does Capable have any intentions of making use of this data on a semantic level?
- Griffin
The similarity you point out comes from the fact that both Capable Networks and RDF are both calling on our collective knowledge of language to solve very different problems. Both efforts are looking for a very simple way to represent a very complex problem. For RDF, it’s a descriptive language for knowledge management in general. For CN, it’s a descriptive language for representing actions observed in a community setting. There are clear knowledge management implications to our system, but we are not somehow using RDF’s methods. Both CN and RDF are leveraging the grammatical rules of the English language to make a difficult problem more manageable.
We do make use of the data on a semantic level. Each sentence in the valuation language has meaning. Each sentence is observed by the CVE, and the meaning is aggregated (interpreted), so that value judgments can be made. The semantics are the core of our system.
Jeff,
With regards to using the data on the semantic level, I wanted to clarify my question.
RDF’s main goal is to facilitate the sharing of data by agreeing on semantics. Meaning, two applications can process the same data just by agreeing on the semantics. My question was really more along these lines. Does CN have any intentions on sharing semantics so that others may use the data in whatever fashion they chose?
If not, I am not sure what the semantics buy you. Without using the semantics to enable automated processing of the data by more than one entity, it’s just like any other internal schema of a system. Though I suppose there is value if there are several internal systems sharing data, but I didn’t think that’s what you described above.
nice, very nice