First IAO workshop notes - Darren

The meeting opened with a specification of scope of Information Entities by Alan, which indicated the involvement of humans or machines to generate or interpret the information. However, there were certain objections to this, because it leaves out several types of information (some examples: the dance of bees can be interpreted by other bees as information describing the location of food; molecular machines use the information within DNA to make copies of it, to make RNA and proteins; perhaps photographs?). It is recognized that what is desired here is actually some sublevel of information. Rather than revise the definition or scope, probably a different term should be used. Alan agrees that the term can be changed. Action item: what term to use?

An assay is a process initiated by an agent (scientist) to measure some quality
 * has_input some material
 * has_output “information entity”

What is the relation between the input material (more specifically, the quality measured) and the information entity? Examples:

information entity	describes	quality of material

1)	110g			relation?	weight of mouse 2)	true			relation? pregnancy state of a woman

here, “describes” is some place holder relation (could also be: about, appears)

For example 1:

Information Entity						Mouse Quality Measurement						“mass quality” of mouse hasValue: number inUnits: unit ofDimension: mass Dimensions Mass has_unit gram, lb, etc.				Length Time

The relation “is_measurement_of” fits well for example one, and in fact allows some kind of error checking to ensure that the units used are appropriate for the quality being measured. Furthermore, it implies what we want it to imply without going too far: that is, there is no statement of veracity, just that it is a measurement. There is general acceptance by the participants. However, the relation is not appropriate for example 2 (or any other type of “classification,” be it binary as in the example or those with more possible bins). Note that to capture some notion of validity/veracity of the measurement, one must know which instrument was used, certain conditions, etc. Action item: what relation is appropriate for example 2? Then, how to merge them into a single relation that works for both?

Is an objective—in and of itself—information? Or is it an objective specification the information? It is possible that an objective is not communicable, but an objective specification always is.