image description as metadata

From: Kari Kraus (KKraus27@AOL.COM)
Date: Sun Nov 12 2000 - 08:27:02 CST

  • Next message: Kari Kraus: "Fwd: image description as metadata" 

    Message-Id: <200011121431.HAA31288@dns.ccit.arizona.edu>
Date:         Sun, 12 Nov 2000 09:27:02 EST
From: Kari Kraus <mailto:KKraus27@AOL.COM>
Subject:      image description as metadata
To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU

    <pre>
    --part1_5e.32ff616.274002b6_boundary Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Content-Language: en

    I'm hoping someone on the list can point me to a study or two on image=20 description as metadata. My sense is that while controlled vocabularies are=
    =20 a favorite talking point among information managers, the free-text=20 descriptions that often accompany visual objects in a database haven't been=20 subjected to the same kind of scrutiny. A representative example that to my=
    =20 mind underscores the need for more conversation:

    In its recently published guidelines on metadata, the Digital Imaging Group=20
    (comprised of representatives from Kodak, Canon, Hewlett-Packard,and Fuji,=20 among a host of others) had this to say: =E2=80=9Conce an image is retrieved=
    , some=20 data that describes the image but is not useful when searching may be=20 included.=C2=A0 For example=E2=80=94=E2=80=98Craig is the guy asleep on the=20= lounge=E2=80=99 is not all=20 that useful when searching, but is useful when describing the content.=E2=
    =80=9D=C2=A0=20

    Now it is not my intention to impugn the recommendations set forth in this=20 ambitious document (a milestone achievement that attempts to establish=20 much-needed jurisdiction over the largely makeshift world of metadata=20 production; its discussion on image capture metadata, for example, is=20 admirably comprehensive and authoritative), but the tautology underlying thi= s=20 quotation (keywords in first sentence: =E2=80=9Cdescribes . . . useful . . .=
    =20 searching=E2=80=9D; keywords in secondsentence: =E2=80=9Cuseful . . . search= ing . . .=20 describing=E2=80=9D) does little to help clarify the nature of the interface=
     between=20 these respective data fields.=C2=A0=20

    Any thoughts on how to define the relationship between descriptive and more=20 strictly classificatory categories of information? Is there a theoretical=20 body of literature in place that looks at the challenge of describing images=
    ?=20
     A set of guidelines or recommendations? Any suggestions would be greatly=20 appreciated.

    Best wishes,

    Kari Kraus University of Rochester mailto:kkru@mail.rochester.edu mailto:kkraus27@aol.com

    --part1_5e.32ff616.274002b6_boundary Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Content-Language: en

    <HTML><FONT SIZE=3D2>I'm hoping someone on the list can point me to a study=
     or two on image <BR>description as metadata. &nbsp;My sense is that while c= ontrolled vocabularies are <BR>a favorite talking point among information ma= nagers, the free-text <BR>descriptions that often accompany visual objects i= n a database haven't been <BR>subjected to the same kind of scrutiny. &nbsp;= A representative example that to my <BR>mind underscores the need for more c= onversation:
    <BR>
    <BR>In its recently published guidelines on metadata, the Digital Imaging Gr= oup <BR>(comprised of representatives from Kodak, Canon, Hewlett-Packard,and=
     Fuji, <BR>among a host of others) had this to say: =E2=80=9Conce an image i= s retrieved, some <BR>data that describes the image but is not useful when s= earching may be <BR>included.=C2=A0 For example=E2=80=94=E2=80=98Craig is th= e guy asleep on the lounge=E2=80=99 is not all <BR>that useful when searchin= g, but is useful when describing the content.=E2=80=9D=C2=A0=20
    <BR>
    <BR>Now it is not my intention to impugn the recommendations set forth in th= is <BR>ambitious document (a milestone achievement that attempts to establis= h <BR>much-needed jurisdiction over the largely makeshift world of metadata=20=
    <BR>production; its discussion on image capture metadata, for example, is <B= R>admirably comprehensive and authoritative), but the tautology underlying t= his <BR>quotation (keywords in first sentence: =E2=80=9Cdescribes . . . usef= ul . . . <BR>searching=E2=80=9D; keywords in secondsentence: =E2=80=9Cuseful=
     . . . searching . . . <BR>describing=E2=80=9D) does little to help clarify=20= the nature of the interface between <BR>these respective data fields.=C2=A0=20
    <BR>
    <BR>Any thoughts on how to define the relationship between descriptive and m= ore <BR>strictly classificatory categories of information? &nbsp;Is there a=20= theoretical <BR>body of literature in place that looks at the challenge of d= escribing images? <BR>&nbsp;A set of guidelines or recommendations? &nbsp;An= y suggestions would be greatly <BR>appreciated.
    <BR>
    <BR>Best wishes,
    <BR>
    <BR>Kari Kraus
    <BR>University of Rochester
    <BR>mailto:kkru@mail.rochester.edu
    <BR>mailto:kkraus27@aol.com</FONT></HTML>

    --part1_5e.32ff616.274002b6_boundary--

    </pre>

    This archive was generated by hypermail 2b29 : Sun Nov 12 2000 - 08:33:53 CST