My name's Mike Bossingham and I normally listen to the list and
keep quiet. However Mark recently talked about stylometric study.
I, too, feel that looking at vocabulary is not a good way of looking
at an author's style. So I developed some software to look at the
word order of sentences.
I looked at word pairs (with every word appearing twice, compared
with the one before and one after). The word is parsed and the
combinations of grammar types counted. The result is then plotted
as a "contour map", giving a finger print for a piece of text.
I have loaded into my website some sample fingerprints for you
to look at. The URL is:-
I am sorry about the quality, but they are scanned out of my thesis - and
thats difficult when it is bound!!
The symbols in the diagrams mean :-
IP - Improper preposition
PA - Particle
SB - Sentence break
CJ - Conjunction
PN - Pronoun
VE - Verb
NO - Noun
AR - Article
PP - Preposition
AJ - Adjective
AV - Adverb
IN - Interjection
A quick look will show that the finger prints look similar where they
should (ie Luke/Acts) and different
where they should.
The software needs a lot more work - for example it only looks at whole
books - and
it needs to be brought into Windows. But I think with some work it could be
a useful tool in this area of work.
I am happy to discuss the techniques used with anyone who is interested and
15 St Mark's Rd.,