3.112 summary of Kurzweil report (54)

Sun, 11 Jun 89 19:26:55 EDT

Humanist Discussion Group, Vol. 3, No. 112. Sunday, 11 Jun 1989.

Date: Thu, 08 Jun 89 22:13:23 ITA
From: Amedeo Quondam / Thomas Walker <G2MFEV42@ICINECA>
Subject: Kurzweil scanning of Italian Renaissance texts

[The following briefly describes the report on the scanning of Italian
Renaissance texts by Kurzweil. As announced in a previous message on
Humanist, this report is available from the file-server, s.v. KURZWEIL
REPORT. My thanks to Cesare Brizio, Amedeo Quondam, Tom Walker, and
their colleagues at the ISR. --W.M.]

The Istituto di Studi Rinascimentali (ISR) of Ferrara (Italy) has
embarked on a project of creating text data bases using various
forms of input including ICR scanning. This has involved an
intensive collaboration with Galileo Centro Studi, an Italian
software and computer services house located in Cento (province
of Ferrara). Dr. Cesare Brizio of Galileo has prepared a fairly
extensive report on the subject of "Recurring errors in ICR
scanning", which describes the text processing phases, along with
the problems encountered in scanning and editing.
Among the questions which Dr. Brizio's report considers are
the following: the consequences of using personnel not
specialized in literature or history; the choice of a final text
format; problems intrinsic to ICR technology (contrast, connected
characters, thin paper); disturbing factors linked to page
structure; the characteristics of a "perfect" page; problems
caused by the specific nature of texts of interest to the ISR
(unreadable accents; characters not available as keyboard keys;
characters not available in the ASCII set such as most Greek
letters, subscripts and superscripts; elimination of line numbers
and note numbers; elimination of non-textual scanned material);
factors of human error (cultural level of the operators;
similarities in character shapes); the scanning/editing procedure
used to process the text, including prior examination of the
material and evaluation techniques; typology and quantification
of errors.
The work done so far has produced encouraging results and
built confidence even in such a problematic environment as
Renaissance Italian poses. Both Galileo and the ISR would
welcome observations, advice and information from anyone
interested in or involved with this field. We would, in
particular, be pleased to launch an exchange of information and
experiences between those concerned with Kurzweil ICR as applied
to the field of Italian Renaissance literature, although
certainly many of the questions raised here are of more general