Vision and Mission
Members
Programme Co-Chair
Messages
Milestones
Programmes
Research
Participation
Job Placements
Admissions
Faculty and Staffs
Students
New Initiatives
Events

 


Research Outline:

Facing intensive global competition, manufacturing companies strive for success. Heavy R&D investment, knowledge engineering and management as well as intelligence of competitors’ R&D activities and technology advancement have become their main concerns. Considering a manufacturing company which specialises in Rapid Prototyping, an advanced technology which has been identified as delivering more competitive products with better quality and shorter fabrication time, accuracy is one of the major issues. With the availability of rich electronic research papers and patents dealing with newly devised and improved accuracy models, an Information Retrieval (IR) system will be a critical tool for R&D engineers to seek inspiration, monitor competitors’ research activities and protect one’s technology know-hows. However, existing IR systems are mainly document retrieval systems based on keyword matching. Furthermore, the knowledge of specific domains, e.g. manufacturing are not currently used by existing IR systems, thereby losing important information and leading to imprecise searches.

Early research on text classification has shown its effectiveness in organising documents according to a predefined structure. In our study, we first focus on the investigation of the possibility and performance of building an IR system by using text classification as the backend engine. In order to harness the merits of manufacturing domain knowledge, we employ a hierarchical form for the text classification. We aim to route the document efficiently while classifying it effectively. We are also interested with the modularisation of such a system to facilitate its application in other domains.

Research Experience:

The first year with the SMA programme gave me the opportunity to extend and enrich my background in manufacturing. I worked together with four other students for an S.M. research project titled “Next Generation Manufacturing (NGM) in Singapore”. This project was co-supervised by Professor Chun Jung-Hoon from MIT and Associate Professor Ngoi Kok Ann from NTU. We were also very privileged to have Professor David Hardt, the IMST Programme Co-Chair from MIT and co-founder of the American NGM project and Professor Andrew Nee, as our external consultants. They shared their valuable insights regarding the future of manufacturing with us.

My current Ph.D. research topic started in my second year. I work closely with the research team under Associate Professor Loh Han Tong. We have biweekly research meetings to share our understanding and regular video conference sessions with Professor Kamal Youcef-Toumi from MIT to discuss my research progress. In addition, I had the chance to discuss on research ideas and methodologies when Professor Kamal Youcef-Toumi and Professor David Hardt were in Singapore. It is a testament to the fact that within SMA, e-mail is not the only mode of communication between faculty and students.

I look forward to my semester stay at MIT as I believe it will be an exciting immersion into the MIT culture. The knowledge and hands-on skills I will gain from the MIT courses will be useful for my research work as well as in fostering more novel ideas. I am passionate to continue my journey with SMA and fulfill my dream of obtaining my Ph.D.

Liu Ying   

 
 
   
   
 Student  : Liu Ying
(Ph.D. 2001/2002 intake - IMST programme)

 Thesis  Advisors
:
Associate Professor Loh Han Tong (NUS)
Associate Professor Tor Shu Beng (NTU)
Professor David E. Hardt (MIT)
Professor Kamal Youcef-Toumi (MIT)

 Research  Title

:

A Hierarchical Text Classification System for
Efficient Manufacturing Information Retrieval
 

 

  < Back   Next