Print

Print


What I was thinking of is how many words does each word follow and/or
precede? I don't think simple frequency counts would be very helpful.

stevo

On Fri, Dec 21, 2012 at 8:10 PM, Leonardo Castro <[log in to unmask]>wrote:

> 2012/12/21 MorphemeAddict <[log in to unmask]>:
> > The books that prompted this line of thought are in Italian and
> Vietnamese
> > (but not in both). Assuming no knowledge of either one, it seems that
> word
> > order and adjacency would be big factors, as well as similarities of word
> > endings (at least for Italian). It seems that the distributional
> properties
> > of various parts of speech would be distinct, at least within a language,
> > maybe for a group of similar languages. E.g., adjectives occur only with
> > nouns or a copula, adverbs with just about everything except bare nouns,
> > articles with nouns, tense particles only with verbs, etc.
>
> Isn't there word frequency analysis technique, similar to what they do
> with letters?:
> http://en.wikipedia.org/wiki/Frequency_analysis
>
> For instance, the word "person" might have approximately the same
> ocurrence frequency in most languages, and maybe the words "house",
> "child", "most-common-animal", "most-common-plant", etc.
>