Background of the NICE Project

1
Background of the NICE Project

• Lori Levin
• Jaime Carbonell
• Alon Lavie
• Ralf Brown

2
Low Cost MT for Minor Languages

• Policy makers can access ideas, viewpoints, and
  information from the developing world.
• Epidemics
• Crop failures
• MT for unforseen translation needs e.g.,
  humanitarian aid.

3
History of the ProjectHow did people from the
United States get involved in indigenous
languages of Latin America?

• Honduras, October 1998 Meeting on information
  technology in multilingual contexts.
• Attended by David Beale and Ruth Connolly (OAS,
  CICAD), Ron Cole (University of Colorado), Gary
  Strong (National Science Foundation, Digital
  Libraries).
• Discussed the digital divide and minority
  languages.

4
History of the Project

• Orlando, Florida, 1999 National Science
  Foundation meeting of principal investigators in
  human-computer interaction.
• Attended by representatives of drug commissions
  from Latin American countries.

5
History of the Project

• Workshop in Manzanillo, Mexico, 1999.
• Attended by funding agencies from Latin American
  countries, and by some scientists.
• Discussed digital divide and digital government.
• Established goals for digital government
• empower people by giving them access to
  government publications
• enable communication between Spanish speaking
  governments and non-Spanish speaking people
• give people access to educational tools
• OAS committed to these goals, with concern about
  drugs and alternative agriculture - growing,
  marketing, etc.
• Report is available.

6
History of the ProjectHow Carnegie Mellon
University got involved.

• Faculty a the Language Technologies Institute
  (LTI) at Carnegie Mellon University (CMU) were
  engaged in research on reducing the development
  time and cost for machine translation for new
  languages.
• Jaime Carbonell (Director of the LTI ) and Gary
  Strong (who moved from NSF to DARPA), made an
  agreement to apply language technologies to the
  goals of the Manzanillo workshop.

7
History of the Project

• Meeting in Uruguay, 1999.
• Attended by scientists from the United States,
  OAS, and funding agencies from Latin American
  countries
• Including Anna Maria Prat, department chair of
  informatics, CONICYT (Chiles national science
  foundation)
• Discussed some possible MT projects
• conducting surveys of health and crops in
  indigenous languages.
• Electronic kiosks with access to government
  information.

8
Current Involvement of OAS

• OAS is committed to the goals of the Manzanillo
  workshop.
• OAS has invited Latin American governments to
  participate in a project for machine translation
  of indigenous languages.
• Colombia responded.
• A meeting took place in Bogota in May 2001.
• OAS does not currently provide money for NICE.
• NICE does not have to adhere strictly to MT
  projects discussed in Uruguay.
• MT projects will be designed in collaboration
  with indigenous communities.

9
CMUs Plans for Establishing Partnerships

• Identify two or three indigenous communities that
  want to work with us.
• Design an MT application that fits in with their
  plans for community development and education.
• Identify scientists who want to work with us
  linguists, computer scientists, etc.
• Identify non-U.S. funding sources for the
  indigenous community and scientists.
• Identify existing programs like ENLACES.

10
Partnership with Chile

• Initial meeting
• May 2000
• Lori Levin, Allen Sears, Rodolfo Vega
• Institute for Indigenous Studies, and Institute
  for Educational Informatics, University of the
  Frontera, Temuco, Chile

11
Partnership with Chile

• Agreement with UFRO
• UFRO provides typing, recording, translation
• CMU provides computers, and salaries on a
  temporary basis

12
Partnership with Chile

• Agreement with the Ministry of Education
  (Mineduc)
• For the year 2001
• Mineduc will support the UFRO team.
• UFRO will record, transcribe, and translate 120
  hours of Mapudungun conversation.