The research proposes to integrate theories and methods from computer-supported collaborative work and learning (CSCW and CSCL) to promote better human-computer intercultural interactions. The research will consist of three phases: 1) identify and categorize intercultural communication problems and determine how they impact group outcomes; 2) apply machine learning to automatically recognize recognize when communication problems arise (or are likely to arise); and, 3) develop and test interventions for improving intercultural communication that could be triggered based on the automatic analysis.

This work has the potential to improve the design of global on-line classroom experiences and business collaborations. At the societal level, the research will help improve intercultural communication and the effectiveness of intercultural teams. This work will also advance the emerging field of automatic collaborative process analysis, which until now has focused only on English-speaking collaborations.

Project Report

New communication tools allow people to interact across national and cultural boundaries in ways that would have been difficult in the past. In virtual organizations, for example, teams of people from across the globe can work together on common problems. However, while multicultural teams benefit from a diversity of perspectives and expertise, they are often challenged by mismatches in social conventions, work styles, power relationships and conversational norms, which can lead to misunderstandings that negatively affect relationships among team members and the quality of group work. The goal of this project was to identify the problems that arise in intercultural communication and develop new computational solutions to prevent or minimize them. Intellectual Merit: This project addressed problems of intercultural communication by (a) conducting laboratory studies to identify and categorize the types of misunderstandings that arise in intercultural dialogues and delineate how these problems impact subjective and objective group outcomes; (b) applying machine learning techniques to coded dialogues with the aim of automatically recognizing when problems arise (or are likely to arise) in an intercultural conversation; and (c) developing and testing interventions to improve intercultural communication that can be triggered by this automatic analysis. The research team spanned three sites: Cornell University, Carnegie Mellon University, and Long Island University. The results of the laboratory studies demonstrated that message content and people's affective and cognitive reactions mutually shape one another on a moment-by-moment basis during team interactions, and that the nature of these mutual influences differed depending on the cultural composition of the group. Participants' negative emotions in intercultural teams had spiraling negative effects on their face-to-face and CMC interactions. The results of the NLP activities led to several new techniques for modeling the dialogues that could lead to new types of automated interventions in intercultural communication. Several new tools were also designed and evaluated, including an interface that provides automated linguistic feedback as a method for reducing problems in intercultural communication and a tool that supplements dialogue with automatically retrieved pictures to clarify the speaker's meaning. Broader Impacts. The project furthered the graduate education of five doctoral students, four of whom have subsequently earned their PHD and obtained excellent faculty or research positions in the field. The project also supported approximately a dozen undergraduate students, and many of these have gone on to graduate study. The project led to numerous peer reviewed publications in prominent journals and conference proceedings, and generated several information resources that will be made freely available to other researchers and interested parties upon request, including experimental materials, NLP techniques, and working prototypes of our tools to improve intercultural communication. Finally, the findings have important implications for society by showing how cultural differences in communication can have negative impacts in intercultural collaborations and by suggesting some strategies to avoid these negative impacts.

Agency
National Science Foundation (NSF)
Institute
Division of Information and Intelligent Systems (IIS)
Application #
0803482
Program Officer
William Bainbridge
Project Start
Project End
Budget Start
2008-12-15
Budget End
2013-11-30
Support Year
Fiscal Year
2008
Total Cost
$676,043
Indirect Cost
Name
Cornell University
Department
Type
DUNS #
City
Ithaca
State
NY
Country
United States
Zip Code
14850