From W3C Wiki
Myth Busting Q&A
There are a lot of misconceptions about the Semantic Web. The goal of this Web page is to dispell those misconceptions. This document should continue to be enhanced as we gain further insight into how various communities view the Semantic Web.
Myth: The syntax is too complicated
Reality: The RDF/XML syntax is not meant for people to read or write directly. It's a computer data format, meant for reading and writing by programs (and hardcore hackers). See AuthoringToolsForRDF and OwlAuthoringTools.
Notes: There was some interest at the last OWLED wksp to include something like "human understandable formats of language primitives" with a potential objective to move towards more standardized way of rendering languages like OWL...
Myth: The logic is too complicated
Reality: Modeling is hard work, usually done by professional software engineers working with help from domain experts (users). Creating ontologies (modeling in OWL) with the appropriate tools is no more complicated than building a UML model or a database schema. See OwlAuthoringTools.
Myth: Global ontologies required
Reality: Patently false. It's all built around using whichever ontologies you need, selected from an open and evolving market of sometimes-competing possibilities.
Myth: It's just a form of syllogistic logic
Reality: Even if that was the only kind of logic on the Semantic Web (it's not), any sort of trivialization of its value, must also discount the value of a relational database (eg: relational logic). If you donÇt think that relational databases have been successful and are valuable, then make Google your query interface for the database, have fun!
Myth: It's not needed in the real world
Reality: It is. Most vendors, integrators, analysts, and standards are adopting RDF/OWL precisely because their customers need its benefits.
Myth: Semantic Web makes you tag everything again
Reality: Natural language with semantic tags is only one way to get data into the Semantic Web, and not a very common one. While some people have reported success with it, most Semantic Web data comes from data sources, not natural language sources.
Myth: Semantic Web won't scale enough to be useful
Reality: Show us a use case where it wont work? In some cases, it's true, performance costs of using Semantic Web technologies will outweigh the gains, but often the increased functionality and east of maintenance and evolution are be the dominant factors.
Myth: Semantic Web is too complex for people to ever understand
Reality: Most of the apparent complexity comes from researchers making sure the basic design is suitable for very demanding applications, not just a stab in the dark. The basic ideas are simple, but they are often explained in the language of certain professional niches.
Myth: Semantic Web is not substantively better than XML
Reality: For mere data interchange it may offer little obvious benefit; for dealing with heterogeneity, data interpretation or incompleteness XML has not got sufficient semantics. Example: EDI works as long as interchange formats are set in advance, if any party changes anything, the templates need to be changed at all endpoints...
Myth: Semantic Web is for academia
Reality: No, but many academics have participated, along with industry, in developing and studying Semantic Web technologies. One doesn't have to be a scientist to use the results of science.
Myth:Semantic Web is top down, whereas Web 2.0 is bottom up (thus better)
Reality: Web 2.0 is actually Top-Down while Semantic Web is Bottom-up. This is becuase The Semantic Web is about the Data & Context (Data Model) side of Web. Web 2.0 on the other hand, is about Presentation and Application Logic (Web Services & APIs). Web 2.0 part is enriched by the Data & Context focus of the Semantic Web. Data combination and resue is natural (meshing) as opposed to forced (mashing) when you add the Semantic Web to the Web Innovation Continuum.
Myth: When the scope of data analysis needs to span years, decades, and centuries, both the scale and provenance capability of a SemWeb infrastructure can be brought into question
Reality: It may be that the current infrastructure is not up to the task, but it looks like the most promising option.
Myth: Nobody is using the Semantic Web
Notes: Would it make sense to say that (almost) everybody is using and relying on (some) semantics on the Web, though... (e.g. reviews relate one person's point with another's; currently the tools apply this semantics implicitly, without the viewer's knowledge). What SemWeb suggests is merely explicitly exposing these relationships and processes, and in turn make them as reusable as data, web content, etc.
Myth: RDF is based on XML
Reality: No, RDF can use XML when it makes sense to do so, but RDF works very well without XML, and people can use Semantic Web technologies quite effectively without knowing anything about XML.
Myth: RDF and XML are competing technologies
Reality: Although users sometimes do have to make a choice between them, RDF and XML generally solve very different problems. Do spreadsheet programs and word processing programs compete? Yes, a user has to chose one or the other for a given task, but in general one is much better suited than the other, so there is little real competition. (This analogy is not arbitrary: spreadsheets, like the Semantic Web and like database technologies, allow the computer to work with the structure of your information. If you're just working with prose, XML and word processors may be better suited to your tast.)