Rewriting the web, page by page
As a semantic copywriter, you get the privilege of rewriting clients' content that ::cough:: 'doesn't need much changing.'
First port of call should always
be NLP - how easy is the content to read? For me, #HemingwayApp
is the tool of choice.
The other data extraction tests:
• taxonomies, relations, lemmatization, keyword prominence
• subsequent social tagging and RDFa
◘ ( #OpenCalais
• and sentiment analyses
are worth Jack if your meaning is hidden beneath rambling sentences.
One word can change the whole meaning of an article in the eyes of a search indexer. Yep, just one word (usually a targeted sentiment or unnecessary adverb) and its placement can totally flip the whole concept.
So when you paste in the original content you have to rewrite and it looks like this - wow. you've got your work cut out.
35 sentences, 2 hard to read, 12 'very hard' to read.
Of the remaining 21 sentences, 4 are part of a bullet list, 9 are headers/sub-headings.
Google and Bing both reported this month having trouble trying to extract data from human-source web pages (KG); it's easy to see why.The long hard road to rewriting the web starts here
Not sure where it leads? Get a guide.
The possibility of not ranking for key terms awaits for those who meander from the path, if those terms are obscured by superfluous forna or buried in the undergrowth. Be clear; be crystal clear!