« Grav (1951-2007) | Main | What's up with Technorati? »

Another nail in the coffin of semantic analysis

undergradNice little test site here although one wonders which undergraduates at which Universities. I tested out a couple of the key blogs I read every day. The boys and girls at Anecdote and Green Chameleon (two collective blogs) are obviously communicating with a wider audience as they came out as High School. Tom Davenport made it to PostGrad level.Then came a shock, Euan can only be understood by a Genius, and I started to wonder if the measure was now one of the degree to which a blog is cryptic in nature. Then it got crazy; I tested out Neurophilosophy which occasionally has me searching for dictionaries and it came out as Junior High School along with Stephen Fry. To cap it all Colin McGinn a philosopher's philosopher if there ever was one only made High School. The whole system gave up with Grandad and declared it itself unable to reach a conclusion, so it obviously has no sense of humour.

Comments (5)

Elmi Bester:

This is another example of growing up in trusting the computing power in cyberspace, without the perceptual context provided by physical space. My husband, who is a Mechanical Engineer, tells the story of a student who was tasked with designing a driving shaft for a motor cycle as an assignment. He presented an impressive report with simulation on the latest and greates packages. The only glitch was that the size of the shaft was more suited to a large truck... and with all the power of computing applied, this common sense check escaped all until the presentation. Back to the drawing board (literally) is getting a new meaning...

Dave, I think you may have misconstrued the function of the test. The "score" is a function of the number of multi-syllable words and long sentences. The higher the level, the more unintelligible the prose. Thus, The Three Little Pigs would get a Grade School level, but so would Hemingway. The lower the threshold, the simpler the language; this is independent of content. I am delighted that both my blogs are rated "Junior High School."

I assumed it was something like that Jay. My point is that the analysis does not produce the measure it claims. Euan's blog for example is very readable ....

Dave, of course you find Euan's blog readable, because you are a wise genius.

How bizarre!

I use semantic analysis to analyse content so as to better understand it from a reductionist perspective in order to inform content classification and creation.

I would hate to ever use it to programatically deduce some form of level of english use.

From my perspective, plain and simple English is always the best use of English.

M

Post a comment




Remember Me?

(you may use HTML tags for style)