Domains of Information

I have a hypothesis that each application/database and stand alone databases are "contained" information domains, and contain sets of data elements that define their context.
If my hypothesis is correct then each of these domains could be described with an ontology, where by the context and definition of the data sets would be shared by the community that is the primary user/consumer of the information within the database.

Each database within a company would have it's own primary ontology, which is nothing more than fully described database schema. Since the guiding rules of relational theory state that you should be at a 3rd Normal rationalization, then the primary ontology is created on this definition of the schema.

Taking this further we can envision that each database is its own domain with its own ontlogy, with the enterprise having sets of ontologies. I am using the term "sets" to mean a collection of ontologies that may be reflective of similar context within combined user communities of interest. This concept provides a method for describing and cataloging information sources across the enterprise. The individual information sources are defined within the context and rules of the primary ontology, meaning that the data within the information source is defined based on the primary community for which the database was developed for.

If we catalog all information sources within a company, and we have formal ontologies for each of these information sources, it becomes evident that we can evaluate the business value of these information sources based on the business value chains that are defined by the organization or company.

So have I sparked any thoughts or comments yet?

I will continue to build this idea in future posts, I would value your input...

What do people think information is?

Thought I would start a series that would begin with a question;
Is the data contained in a database finite, contained, and controlled?

If you assume that a relational database is supporting an application, and the application uses the database to store, retrieve, update, and query the application data.

Would you believe that the data is contained to the data definitions that are used by the application and it's users? Also would the application control the data definitions and not allow data that was not defined by the application? Would you also agree that data is finite within the boundaries of what the application has specified in its domain of understanding?

Can I make a statement about the applications database, the data within the database is contained within a domain that is defined by the applications semantic definition of data entities and tables, and the relationships between the data sets that make up the database.

In other words, the data within a database is composed of sets of data elements that are grouped together based on the definition of a vocabulary, (i.e. taxonomy) which is understandable by the application. And once defined, can only be changed if the application is changed to accept the new definition or new element.

Within an organization there may be many applications that have relational data sources which support the applications. If each of these applications was developed by and for different users, then the vocabulary of those users would be reflective of the data set definitions within those applications databases. So the semantic meaning of the data sets within these databases would only be relevant to those users who know and understand their vocabulary.

Should you force all users to accept only one vocabulary within a company or organization? Is this how an organization really executes, or is a company composed on many diverse communities that perform functions based on their communities vocabulary?

So just some random questions I will post more on this but I am looking for your reaction to the points I raised...

Information Architecture is not a Web Site Architecture

Before you can manage enterprise information you must first architect enterprise information. Our industry has relegated the meaning of Information Architecture to mean Web and Portal site content, document management, business intelligence, business analytics, and content management, which is only a small fraction of enterprise information. We need to resurrect the true meaning of Enterprise Information Architecture and elevate it to a more important role in the Enterprise Architecture realm.

Information Architecture should and must mean the creation and design of information structures that support business function and business users within their context of usage and meaning. An Information Architect must identify the data elements within data sources that are needed by specific users and processes throughout the business functional lifecycle, and semantically align these data sets to the context of the user and process that needs this information. Information Architecture is driven by the users meaning and context not just the business process context.

Information Architecture is more than Data Architecture and the two are distinctly different. Information is the meaning and context of data, in databases today (whether structured or unstructured) the meaning and context of data is not stored with the data element. Data tables and data records only store the structure and syntax of the data not the meaning.

A database is designed and implemented from the requirements and definition of entities that will be stored and manipulated in a physical database. The meaning of the data is reflected in the structures and relationships between tables and records that a group of designers create. Each entity must have one and only one meaning or you break the fundamental theory of data design. The intent of the database is to store, manipulate, and report on the data that the organization needs to accomplish its business for a specific set of users and processes. If you use this data for any other group of users, that the database was not designed for, then that group of users and processes must either accept the meaning of the data or translate that data into meaningful data for their use.

It is evident that business users and process need “their” own data, just look at the hundreds and thousands of databases that a company maintains. There is not “one version” of truth when it comes to the information a company needs to be effective, efficient, and profitable in today markets. Forcing everyone in an organization to use one defined set of information is like forcing everyone on earth to speak one language. But our industry continues to force this definition by addressing the problem with Data Warehouses, Data Marts, Metadata, MDM, and Enterprise Data Models, which all have failed in meeting the business needs of an enterprise because it is focused at the data not the information needs of the users. When I say the word “ship”, do you think of something on the water, or do you think of moving a package from one point to another. Without context and meaning “ship” could mean several things to a user. This same principle applies to the data that is stored in databases today.

Just getting started

Just getting started with this new blog, I plan on writing about what I find of interest in my career and clients that I have had.

So bear with me and more will come as I have the time to share my thoughts and interests