Ontolog Discussion: Tagging the Ontolog Forum for Matching Interests within Compatible Schedules

Proposed Ontolog project - Nicolas Rouquette / 2005.06.23-12:39 PST

Proposed Charter

Background

Thanks to PeterYim's efforts, the "CIM" collaborative-work-environment ("cwe") is host to many interesting forums that share many viewpoints on the broad topic of "ontology". During a recent conference call, http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2005_06_23, Brand Niemann reported on a number of potentially very interesting meetings, reviews, conferences where various state-of-the-art techonologies and concepts will be presented and discussed; for example, at the upcoming Extreme Markup conference: http://www.extrememarkup.com/extreme/

This presents a practical dilema for interested members within Ontolog community, the SICoP and other related forums: how much time to spend googling and swoogling (there's such a thing as "Swoogle", a kind of semantic version of google) for references that might lead to useful information. Brand Niemann also reported on recent interest within the CIO council http://www.cio.gov/index.cfm for reverse engineering an ontology directly from software.

In the spirit of the cobbler's family having the worst shoes in the community, Ontolog and others involved in the broader context of the "SemanticBlogosphere" lack a practice of using Semantic Web technology for tagging, annotating, indexing, searching and populating the forums in a way that allows forum practitioners to be the early adopters and early beneficiaries of state-of-the-art advances in semantic web technology.

Within the "SemanticBlogosphere", the genie of semantic tagging and annotation for ontological construction and query/search support is out of the bottle: while wanton ontological extraction is, in some sense, technologically feasible and, in principle, practical with sophisticated tools, it quickly leads to a number of ontology management issues with respect to complex issues of information management, information flow across various forums, heterogeneity from different viewpoints among forum members and the broader public community.

The above issues require addressing the notion of tagging the Ontolog-cwe itself and others in the larger context of integrating multiple sources of semantic information. From the specific charter of Ontolog, such efforts raise further issues of vendor neutrality, openness and practical interoperability among whatever tools and infrastructures involved in migrating, say, the Ontolog wiki to a semantically-enhanced Wiki.

What's next?

First, like any good project, this one would need good requirements. Writing requirements in wiki pages in text would clearly be a sign of severe misery in the Ontolog Forum cobbler's membership.

Second, a meaningful discussion about organizing this project would require, at minimum, some kind of informed review of the current available options to decide on a realistic strategy for implementing it.

There are different strategies possible:

Incremental step-by-step approaches
Semantic mining
Transformation

These are but a few options among those discussed in recent conferences, e.g., http://www.dagstuhl.de/04391/Materials/

Incremental approaches

e.g:

Add support for annotation-capable browsers (e.g., http://www.mindswap.org/2004/SWOOP/) so to support real-time, multiple-writers annotation of Ontolog Forum discussions.

Add standard ontologies for tagging, e.g.: DC, FOAF, DOAP, ...

At this point, it might be easy to either design or extract an "OntologForum" ontology that would be particularly tuned for the needs of the forum.

e.g., DC, FOAF, DOAP tagging

Semantic Mining

e.g.:

Using SUMO, WordNet and other "standard" dictionaries with semantic annotations, reverse engineer a semantic view of the current forum content

Extract/refine an Ontolog Forum ontology from this semantic view and use that ontology for annotation of future pages using an annotation-capable browser for editing the forum's pages

Transformation

e.g.: the current Ontolog Forum is viewed as "version 1" of a complex system and we are embarking on building "version 2"

Define a new OntologSemanticForum system with a combination of ontologies, tags, etc...

Using mining algorithms and transformation engines (http://www.programtransformation.org), extract the information we care to preserve from "version 1" and import it into "version 2"

Switch to the new OntologSemanticForum

Archive and retire the old Ontolog Forum as a read-only relic of our cobbled past

Comments / Suggestions

... (please add comments/suggestions here or, if you have the inclination and disposition, edit the description above) ...

Semantic 3D space
- Ontology is not enough to add semantic to the web. Semantic deals with meaning and meaning has to be considered in a cultural context. So, defining a dictionary of terms and an ontology to relate terms is not enough if we cannot define the culture to relate to the ontology. See Beyond tagging. (-- Dario de Judicibus / 2008.03.01)