-----------------------------------------------------------------------
Speaker:        Dr. Cheng-Jiun YUAN
                Department of Computer Science
                University of Kentucky

Title:          "ELINK -- A Flexible, Semi-Automatic System for Linking
                 Texts Between Manuscript Images and Transcription

Date:           Monday, 12 May 2003

Time:           2:30pm - 3:30pm

Venue:          Room 3006 (Phase I, via lift no. 3)
                HKUST

Abstract:

ELINK is a semi-automatic system for linking the corresponding texts
between the manuscript images and the transcription - a textual version of
the manuscript.

At a minimum, a digital edition of a manuscript consists of the digitized
manuscript images and its transcription. Having the corresponding texts
linked between the manuscript images and the transcription is very desired
because it enables the users of the digital edition to find the
corresponding text quickly and enjoy other benefits provided by both data
sets. Unfortunately, manually establishing correspondence for every letter
in the manuscript is a difficult and labor-intensive editorial task, and
fully automated solution similar to optical character recognition (OCR) is
unlikely because the manuscript images are noisy and unconstrained.

To address the problem, I design the ELINK with the goal to

1.      maximize automated performance by exploiting additional cues, such
        as structural information of the texts from the transcription and
user's
        feedbacks

2.      minimize the user's effort through carefully designed user
        interfaces and user interactions

3.      maximize reusability and portability by designing the system as an
        extensible and configurable software tool (i.e. component based)



******************
Biography:

Dr. Yuan received his B.Sc and Ph.D in Computer Science at the University
of Kentucky, USA in 1993 and 2003 respectively. He has been actively
involved in two digital library projects, the Electronic Beowulf and the
Digital Atheneum, at the University of Kentucky. His main research
interest includes digital library, hypermedia authoring, Internet
technology, image compression, image recognition, and document image
analysis & character recognition.


                For enquiry, please call 2358 7008
                        ** All are Welcome **
--------------------------------------------------------------------------