From RDFaWiki

Using an RDFa Extractor to Check Your Work

It is considered best-practice to use an RDFa extractor, such as Fuzzbot, to check your work when authoring RDFa. XHTML pages can get complicated, so it becomes hard to tell exactly what sort of triples you are generating without using an extractor of some kind to check your work.

Reasoning

Like HTML, it can be easy to make mistakes that have unintended consequences when authoring an XHTML+RDFa document. Typically, the tools that you use to author content have options to check the validity of your input.

While errors in markup with XHTML can lead to display errors, it is doubly important to check your RDFa markup because mistakes made in RDFa lead to logical errors. These logical errors can be far more harmful than display errors because they can cause errors in computer reasoning. For example, if you mark up a song that you created using the Audio RDF vocabulary and specified the license as Creative Commons when you meant to use standard copyright, someone might think that they can distribute the song to a peer-to-peer network without your permission.

Bad RDFa data can also hurt your search engine ranking, so make sure to check your RDFa triples before publishing your pages.