It has been fun.
One of the things we have comparing is the difference between word counts about web sites and the semantic (important concepts) in a web site.
To show what I mean the next two extracts are a word count and a concept count of what I have said about research and evaluation on this blog.
This is a word count:
able advertising applications ave blog comments company content conversation david editorial ethics evaluation google identify internet labels links management measure media news number obama online organisation pages people per phillips pm post pr practice public relations relationship research results search semantic significant site social twitter value view web work yahoo
created at TagCrowd.com
and this is the semantic view:
conversations deadlines dimension edited evaluation geography journalists label media morals morphs newsletter opinion page philip platforms posts pr priorities professionals public relations research showing tools version young
created at TagCrowd.com
The difference is huge.
The word count shows words that are common in the discourse while the semantic view is about meaning and the drivers of my posts.
Of course, there is a role for both forms of analysis but by far an away the most informative is the semantic analysis.
In bigger corpora my experience is that word counts become ever less helpful and semantic analysis offers real insights.
At Blecom, Bruno and I showed this form of analysis as a proof of concept for some pretty big networks (in real time too) and the results were very interesting.
No comments:
Post a Comment