权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Development of knowledge-creation multilingual collaboration tool

知识创造多语言协作工具开发

基本信息

批准号：
19300036
负责人：
YOSHINO Takashi
金额：
$ 12.06万
依托单位：
Wakayama University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (B)
财政年份：
2007
资助国家：
日本
起止时间：
2007 至 2009
项目状态：
已结题

项目摘要

The purpose of this research is to establish multilingual user participatory methodology to produce new knowledge among users who use different languages on multilingual collaboration environment. The following conclusions can be drawn from the results of the research.(1) We evaluated the effect of the translation repairusing back translation. We evaluated how accuracy result of translation repairs when people rewrote an original sentence.(2) We performed the experiment of translation repair to evaluate the effectiveness of providing inaccurately translated segments. Providing inaccurately translated-segmentscannot contribute to the reduction of the translation repair costs. Therefore, providing inaccurately translated-segments would be not effective to support the translation repair.(3) We evaluated the interpersonal acceptable response time in real-time remote text-based communication. We discussed the duration for whicha user can wait for the other user's response. We verified theeffects of the providing of input status and the elapsed time of communication.(4) We proposed a method that reduces the inaccuracy of judgment. We verified the efficiency of the proposed method in a chat communication using the Wizard of Oz method. Our method can apply to the communication that is requires accuracy.(5) We have developed an All-for-one type multilingual conference support system which supports contents understanding to compile content of a conference by Japanese participants. We performedan evaluation experiment using our system, we confirmed the usefulness of our proposed functions.

本研究旨在建立多语种使用者参与式方法，让使用不同语种的使用者在多语种协作环境中产生新的知识。从研究结果可以得出以下结论：(1)对回译修复翻译的效果进行了评价。我们评估了当人们重写一个原始句子时，翻译修复的准确性结果如何。(2)我们进行了翻译修复实验，以评估提供不准确翻译片段的效果。提供不准确的翻译片段可能无助于降低翻译修复成本。因此，提供不准确的翻译片段将不能有效地支持翻译修复。(3)我们评估了实时远程文本交流中的人际可接受响应时间。我们讨论了一个用户可以等待另一个用户响应的持续时间。验证了输入状态的提供和通信时间的影响。(4)提出了一种减少判断误差的方法。利用Oz向导方法验证了该方法在聊天交流中的有效性。我们开发的All-for-One类型的多语言会议支持系统支持日方与会者对会议内容的理解。我们利用我们的系统进行了评估实验，证实了我们所提出的功能的有效性。