RepresentsUseCase

From AI: [

1. You have the same data in more than one "format", e.g. XML, or Tab delimited.
2. To relate two report graphs generated from the same data (visualization)
3. hypothesis. First english is presented, second logical is presented. ]

I think that all points above are about the relation between an information content entity and some form of representation: format, graph, linguistic. We use the model proposed by Riichiro Mizoguchi (I am sure Alan knows about this model).

In this model there are three essential entities:

1. information content entity

2. representation ( we have subclasses: logical representation, textual representation, mathematical representation)

3. material entity

4. and 3) are covered by IAO and BFO wonderfully. (MC: For example, relation between information content entity and material document is that the material document bears a specifi)ally dependent continuant that is the concretization of the information content entity.] - I completely agree with this.

5. information content entity can have multiple representations. I am struggling with mapping of representations to OBI and IAO classes. The nearest is information carrier (SDC), the definition fits well, but I am not completely sure given the examples of the usage (The pattern in ink on a page of a paperback novel]). What is a relation between an information carrier and the corresponding information content entity? If assume that the mapping is representation = information carrier, than the relation is "represents". This is a fragment from a real use case:

---

Systems Biology models analysis use case study of iND750 systems biology model has_proposition h1 h1 is_instance_of hypothesis

h1 is_represented_as "iND750 systems biology model prediction are not consistent with the results of the investigation into the reuse of formalized experiment information for rich media"

"iND750 systems biology model prediction are not consistent with the results of the investigation into the reuse of formalized experiment information for rich media" is_instance_of text representation

h1 is_represented_as not_match(model_iND750_prediction,re-use_study_results,rich_media) not_match(model_iND750_prediction,re-use_study_results,rich_media) is_instance_of logical representation

study of iND750 systems biology model has_proposition h2 h2 is_instance_of hypothesis

h2 is_represented_as "iND750 systems biology model prediction are not consistent with the results of the investigation into the reuse of formalized experiment information for minimal media"

"iND750 systems biology model prediction are not consistent with the results of the investigation into the reuse of formalized experiment information for minimal media" is_instance_of text representation

h2 is_represented_as not_match(model_iND750_prediction,re-use_study_results,minimal_media) not_match(model_iND750_prediction,re-use_study_results,minimal_media) is_instance_of logical representation

---

This is a simple example, in other studies we have thousands of hypotheses and an efficient representation is truly essential for processing. I do not expect from OBI to cover such cases as it is still not typical for biological experiments, but IAO should be able to support such seriously information-driven investigations.

There is another point to discuss: Is it possible to have a relation between a process and information content entity (we are currently using has_proposition). An actor must have some information in order to initiate the process; the process is governed by this information. Example: investigation has_proposition hypothesis; investigation has_proposition goal. Of course I can use plan specification has_part hypothesis/ goal But is this enough? I think that a direct relation between a process and information content entity would be useful. But I can live without that.