we need a common metadata model for describing Open Government Data apps
I am very excited about the recent announcement of Facebook: http://developers.facebook.com/docs/api . It is very close to what I imagined writing the linkedPIM specification. I have some remarks though:
- semantics still need to be better captured, using JSON does not solve anything – I would use an OWL model to define the standard
- the model will need to be extended, another argument for using RDF/OWL
- Facebook seems to be preparing for competing with linkedin, which is very natural imho I base this assumption on example dataset:
- profile model is really poor, it should be much more like in the address book (telephones etc) – base it on the vcard
- they still try to keep the social web in frame of their own portal. it would be better to claim ownership of a protocol and semantic model which every portal can use (become the main layer in the service architecture)
"name": "Director, Product"
P.S. The old “Chinese” curse is fulfilled. We are living in exciting times.
Note1 – the scope of these definitions is limited to semantic web discipline.
Note2 – 05.10.2009 – updated thanks to a comment from Horst.
Note3 – 06.10.2009 – changed “inference” to an effect instead of process/action (thanks Horst as well). If it is an action, then it is part of the reasoning action. And reasoning action is part of the inference action. So, inference=reasoning, which is not ok.
There is a bunch of definitions about each of the terms, here I try to define with the purpose of marking the main differences and relations between “inference”, “reasoning” and “consistency check”.
inference – the effect of triples added to a model achieved via reasoning (using OWL / SWRL rules / etc.)
reasoning – process of calculating logical consequences from a set of asserted facts or axioms (from wiki). The goal of reasoning can be an inference or a consistency check.
consistency check – verify if there are no logical clashes (contradictions) in the given model
I was for quite some time thinking on how to ask for my son (3 yrs old) what did he dream during the night and get an answer.
The problem being the concept “dream”. He has to know what it means to answer the question.
Today, getting him to sleep, he said he is afraid. When I asked why, he said “that bad cartoon will start”.
What can I say – TV influence, surely was not how it was before some 100 years.
Well, at least it’s good that he didn’t say “that terrible youtube movie will start” or something like that.
For that, still some things must change.
This is a post to test the facebook integration.
If it works, done with wordbook.
Going to i-semantics + i-know conference 2nd-4th of September.
My current schedule (based on available information, might change when the info changes):
- Towards a New Generation of Social Networks: Merging Social Web with Semantic Web
- Semantische Suche in Medienarchiven (praxis)
- Applying Organizational Learning to Enterprise Knowledge Maturing
- Understanding Maturity Models. Results of a Structured Content Analysis
- A Knowledge Workbench for Software Development
- The European Dictionary of Skills and Competences (DISCO): an Example of Usage Scenarios for Ontologies
- Web x.0 – Was bringt wirklich neuen Nutzen für Unternehmen? (praxis)
- Semantic Web Awareness Barometer 2009 – Comparing Research – and Application-oriented Approaches to Social Software and the Semantic Web
- Maturity and Applicability Assessment of Semantic Web Technologies
- Can Software Developers Use Linked Data Vocabulary?
- Monitoring RSS Feeds
- Wissensmanagement im Projektmanagement (praxis)
- Der Einsatz von Web 2.0 in der Internen Kommunikation (praxis)
- Ontology Change Management
- SALERO Intelligent Media Annotation & Search
- Best Practice Forum
- Die dunklen Seiten des Wissensmanagement und Web 2.0 (praxis)
See you there 😉
Don’t ever try to explain to a woman how modeling in OWL is done
To explain how that works, you will typically try model her to represent the world she lives in, and then you would have to reveal:
1. the modeling would typically start with defining her an owl:Thing
2. even if you don’t use a class, she would complain being someone’s property
3. owl:Nothing won’t make her happy, either