The Object-Oriented Content Transmission Metamodel as a Metamodel for Information Transmission

In Figure 5.1 we illustrated the main components of our Object-Oriented Content Transmission Metamodel, which were furthered develop in the following sections. By comparing this metamodel to the one just described by Shannon and Weaver we can highlight the following differences:

In our model, the stream to be transmitted is no longer seen as a stream of bits with no abstract meaning, information is an abstraction of the actual content, in other words, a stream of meaning. We are therefore not interested in removing data redundancy as such but rather on transmitting the appropriate meaning that the source carries within. And what is appropriate is, as in any other kind of model, related to the context, application or what the receiver is expecting to receive and able to interpret. In a general situation, we do not worry about fidelity to the original data but rather to the original meaning.

In this sense, noise is thought of as anything added to the original piece of information that is likely to change its meaning or make it difficult to understand. Thus, the traditional definition for noise as a change in the bitstream being transmitted, would only fit our definition if the change is substantial and can produce a change of meaning.

Our channel does not transmit encoded data as such but rather extracted content description or meaning. Because of this Shannon's laws of channel capacity do not hold true and we are able to well exceed its theoretical limits.

On the other hand, and as Cockburn points out [Cockburn, 2002], S&W's metamodel considers a constrained channel while in human communication the channel is unconstrained and the success of the information transmission process depends on the shared experience between the transmitter and the receiver, which enables them to use a common vocabulary. In our metamodel, the success of the process is also based on the shared experiences or knowledge between the transmitter and the receiver. As a matter of fact this is only constrain.

In order to accomplish all this the traditional S&W metamodel has to enhanced and particularly two new blocks have to be added.

On one hand, in the transmitter we need to analyze and extract content or meaning from the original data source. This process was explained in section 5.2.1. On the other hand, the transmitted message has to be interpreted and rendered into new data in the receiver. This process was also explained in section 5.2.2.

2004-10-18