Issues relating to Database and Ontology

Please add new thoughts and remarks below. Date and identify yourself first. Do not edit others' work (unless its clearly a typo), and annotate it by a idented bullet right below the paragraph you are commenting on (again, identify yourself and date the comment.)

• From Tatiana Malyuta / 2006.08.18 - see: http://ontolog.cim3.net/forum//ontolog-forum/2006-08/msg00076.html
  • Hello, I am the person who requested the discussion of ontology-database topic.

Thanks for responding. I tried to analyze the opinions, added my own concerns based on extensive data modeling and database implementation experience and very limited experience in modeling ontologies, and suggested the list of problems that I consider interesting for discussions (please see the attached file). Maybe this list can be helpful for defining the topics and structure the discussion about onotlogies/databases in the mini series proposed by Peter?

• • The following is the attachment in the message:

Questions of a data/database person about ontologies

I agree with the opinion that the Semantic Web/ontologies should be the glue holding together information/data assets:

Data component of an Information System or the Web is not homogenous in terms of purpose, role, ways of access, etc. Not everything looks as ontology to me and not everything is currently implemented as ontologies; I see several data layers: data, metadata, and ontological data. For these data components to fulfill their purpose and role in the most efficient way, they have to be presented through different models and implemented with the help of different technologies.

We have huge amounts of data accumulated in various data sources already. Semantic Web was intended to be a web of connections between applications and data sources.

• • Therefore, the questions.
  • 1.How to make different "data" parts work together?
    • Ontological community is very involved in the problems of merging (or integrating, or federating, etc.) ontologies. In addition to the problems of "horizontal" merging at the ontological level, discuss how to merge ontologies "vertically" with metadata and data. Sub-questions in reference to ontologies-databases issues:
    • a.How to utilize limited semantics of existing databases for building ontologies?
    • b.How to utilize ontologies for building new databases?
    • c.How to organize interfaces between ontologies and databases?
  • 2.What methodology of ontological data design has to concentrate on?
    • In many cases methodology of design of ontologies is discussed on a very general level in terms of the life cycle. If not general methodology of conceptualization, then maybe recommendations for conceptualization can be discussed to provide the designers with guidelines for design of flexible, extensible, consistent ontologies.
  • 3.How to implement and manage ontologies?
    • Maybe not all problems of data management are important in ontologies management, e.g. concurrent access and transaction management. However, performance issues seem to be important. Sub-questions:
    • a.What are the problems of ontologies management?
    • b.What are the approaches to resolve the problems of ontologies management?