I have a lot of errors when I want to load DBpedia dataset. For example:
22007 XM003: XML parser detected an error: ERROR : Tag nesting
error: name 'img' of end tag does not match the name 'p' of start tag
at line 4 column 432 at line 4 column 438 of source text
04/02/skos/core#" xmlns:xsd="http://www.w3.org/2001/XMLSchema#"></img></p>
----------------------------------------------------------------------^
Ok, let's find the line where the error occured (I put a line break, so it is easier to see):
<http://core-project.kmi.open.ac.uk/data-description> <http://purl.org/rss/1.0/modules/content/encoded> "<h2 xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\">What data are exposed</h2>\n<p xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\">The CORE project exposes data about the aggregated content. The following schema shows the kind of metadata CORE holds about each resource. </p>\n<h2 xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\">Data Schema</h2>\n<p xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\"></img></p>
\n<h2 xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\">Data License</h2>\n<p xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\" xmlns:dc=\"http://purl.org/dc/terms/\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:og=\"http://ogp.me/ns#\" xmlns:rdfs=\"http://www.w3.org/2000/01/rdf-schema#\" xmlns:sioc=\"http://rdfs.org/sioc/ns#\" xmlns:sioct=\"http://rdfs.org/sioc/types#\" xmlns:skos=\"http://www.w3.org/2004/02/skos/core#\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema#\">All data from CORE (unless otherwise specified) are available under the a Creative Commons Attribution 3.0 Unported License. </p>\n"^^<http://www.w3.org/1999/02/22-rdf-syntax-ns#XMLLiteral> .
Why the heck Virtuoso tries to check HTML/XML tags consistency inside the literals?! Is it possible to turn it off? I have too many errors in the dataset, it is a waste of time trying to find all lines with errors and remove them by hands.
Your main question is why Virtuoso tries to check whether the lexical form you've shown us can be parsed as correct XML content. Virtuoso does this because it tries to conform to the RDF 1.1 specification.
According to that specification, if the datatype of an RDF literal is
rdf:XMLLiteral, then the lexical form should be "well-balanced, self-contained XML content" (per XML 1.0).