osherenko at gmx.de
Mon Oct 9 13:13:00 CEST 2006
> 1. attributing any given poem to a particular style is likely to be
> highly subjective, and controversial among scholars of poetry
It doesn't matter much. The category doesn't have to be objective, it must only be in some way comprehensible for a human.
> 2. what taxonomy of types of 'poetic style' would anyone use? Are there
> agreed types and terms?
I see, but there is always some way that is usually specified as a particular trend in poetry and real experts can always determine immediately the literary trend or even the poet.
> 3. what level of granularity of categorisation should be used? Are there
> sub-types ("late romantic")? Is there a hierarchy that can encompass
> all types and sub-types? How many levels does it need to have?
It would be very nice if I had such information since I can always throw away the unnecessary information and leave information I need. In your example, I assume you can always consider "late romantic" as simply "romantic". Such transition ("romantic" to "late romantic") is probably not possible because you would need some additional information e.g. about the authoring time.
> 4. what do you do with poems that fit into more than one style, or are
> at a boundary between styles?
If there is more than one similar poems of this kind, I can define a special mixed category or if it is not appropriate must decide for a particular category loosing some information. Otherwise I don't study this poem in my corpus.
> 5. do the categories fit for poetry from different languages, countries
> or traditions, and for translations?
For the sake of simplicity I work only with one language. Of course, I can also use translations as long as a particular translator uses words and phrases of the destination language that are sufficient to understand the style of the poem.
> Now, it may be that these problems can be raised with any form of
> categorisation. The way to address these sorts of problems is to find
> measures which are consistent, transparent and as objective as possible.
No. They simply have to be consistent.
> I'd be very interested to know more about why you want this though.
> There may be other ways to approach your research question.
I could imagine you would be rather disappointed. It is surely a "sacrileg" and I don't want to be cynical but I want to test my approach to opinion mining (emotion mining) for categorizing poems.
> Best wishes,
> Alexander Osherenko wrote:
> > It's probably an incorrect word 'annotation'. Actually it's important to
> know that a particular poem is said to be a representative of a particular
> style. It is not enough to assume that the poems of a particular author
> are all of some style since humans and also poets do change their authoring
> style during the life.
> > -------- Original-Nachricht --------
> > Datum: Mon, 09 Oct 2006 10:11:17 +0100
> > Von: Martin Wynne <martin.wynne at oucs.ox.ac.uk>
> > An: Alexander Osherenko <osherenko at gmx.de>
> > Betreff: Re: [Corpora-List] Poetry
> >> What do you mean by annotation?
> >> Alexander Osherenko wrote:
> >>> Hello!
> >>> Has anybody seen or maybe heard of an annotated poetry corpus? For
> >> example, poem text and its annotation (romantic whatsoever). The
> >> project is very interesting, but unfortunately without annotations.
> >>> Best poetic wishes
> >>> Alexander
> >> --
> >> Martin Wynne
> >> Head of the Oxford Text Archive and
> >> AHDS Literature, Languages and Linguistics
> >> Oxford University Computing Services
> >> 13 Banbury Road
> >> Oxford
> >> UK - OX2 6NN
> >> Tel: +44 1865 283299
> >> Fax: +44 1865 273275
> >> martin.wynne at oucs.ox.ac.uk
> Martin Wynne
> Head of the Oxford Text Archive and
> AHDS Literature, Languages and Linguistics
> Oxford University Computing Services
> 13 Banbury Road
> UK - OX2 6NN
> Tel: +44 1865 283299
> Fax: +44 1865 273275
> martin.wynne at oucs.ox.ac.uk
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
More information about the Corpora-archive