 |
|
Information Mining for Digital Libraries Project Wins HP
Grant
|
|
Associate Professor Tan Chew Lim (in blue) discussing recognition
of text from curved
surfaces with his graduate
students. |
|
|
An NUS School of Computing
(SoC) research project aimed at easing the burden of
information retrieval for educationists and students has
clinched a prestigious grant. Worth more than US$58,000 in
total value, the grant is given by HP to higher education
institutions to promote innovations in teaching and
learning. Only 14 institutions from around the world have
been selected in 2006 to receive the HP Digital Publishing
for University Teaching and Learning grant, which provides
each selected institution with over US$43,000 in HP
technology and US$15,000 in cash.
Associate Professor Tan Chew Lim, Director of the Centre for
Information Mining and Extraction (CHIME) in SoC, is the
Principal Investigator (PI) for the project. Entitled
‘Document Information Mining for Digital Libraries’, the
project is expected to generate three main outcomes over 18
months: new techniques on document image analysis and
information search, a document image search engine, and a
full fledged educational resource portal.
Associate Professor Tan is actively involved in research on
methods for improving information retrieval from document
images. For instance, current text recognition software
requires that text be scanned from a flat page. Problems
arise when a page from a thick book is to be scanned as the
book cannot be opened flat because of the thick spine. The
text nearer the spine would then appear on a curved rather
than flat surface, and the text on the curved portions would
then be ‘misread’ by the software. Associate Professor Tan
and his research students devise solutions such as
techniques that take the curvature of portions of the page
near the spine into consideration, and pre-process the
affected texts before they are handled by standard text
recognition methods.
Associate Professor Tan’s research efforts in document image
analysis and recognition techniques are complemented by
those of his co-PI, Assistant Professor Kan Min-Yen. Dr
Kan’s efforts centre on building and enhancing a document
image search engine. He has led efforts to set up mirrors of
the CiteSeer computer science digital library in SoC. The
library consists of over 700,000 scholarly documents, which
are used extensively by students and faculty alike. Dr Kan
currently teaches a course on Digital Libraries, targeted at
advanced undergraduates and beginning graduate students who
wish to understand real-world issues in building, using and
maintaining large volumes of information in digital
libraries. His involvement and findings in the project would
enrich the learning of the students. The grant equipment
would also help with the course in coming up with novel
methodologies for learning and disseminating information.
|
|
Associate Professor Tan Chew Lim (centre) shares a light
moment with
Dr Kan Min Yen (right) and Associate Professor Danny Poo (left). |
|
|
Project collaborator
Associate Professor Danny Poo takes care of the third aspect
of the project – developing a full fledged educational
resource portal. He is spearheading a large-scale project on
building a repository of websites and structured contents
for the Singapore education sector. Towards that end, he has
already set up an Education Taxonomy Portal, or E-TAP, which
helps school teachers and students and students navigate and
look for information that they need more efficiently.
Further details on E-TAP may be found at:
http://www.comp.nus.edu.sg/~dpoo/res_etap.html. The
CHIME website is at
http://www.comp.nus.edu.sg/labs/chime/ (for CHIME), and
the CiteSeer site is at:
http://citeseer.comp.nus.edu.sg/cs.
The 2006 HP Digital Publishing for Teaching and Learning
grant is awarding a total of some US$812,000 to 14
universities in nine countries. From 2004 to 2006, HP has
committed more than US$36 million in grants to more than 650
schools worldwide to support HP’s broader education goal of
transforming teaching and learning through the integration
of technology. “HP believes digital publishing technology
plays an important role in transforming teaching and
learning,” said Bess Stephens, Vice President, Philanthropy
and Education, HP. “Faculty members will benefit from the
technology by using it to tailor content to the individual
student; students will benefit by receiving rich-media
content not found in traditional textbooks.” More
information about HP’s philanthropy efforts is available at
http://www.hp.com/go/grants.
|
