Doctoral Consortium for Students in AI; July 28, 1997; Providence, RI

学习如何做研究最重要的方面之一是与经验丰富的导师就研究计划的各个方面进行早期和详细的讨论。不幸的是，大多数学生在他们的论文的初始阶段得到的反馈很少，除了他们的直接导师。参加会议是认识其他研究人员和讨论工作的好方法。然而，一个学生在他/她工作的早期阶段得到机会与成熟的研究人员讨论他们的初步想法的机会不是很好。美国人工智能协会会议上的博士联盟（DC）计划改善这种情况。即将开始论文工作的学生被挑选出来，在一群精心挑选的教师小组成员面前展示他们的研究计划。与传统的会议演讲不同，DC的大部分时间都用于学生和教师小组成员之间的讨论。每个学生还被分配了一位教师导师，他会在更详细的层面上与学生合作。学生既可以从自己的演讲中学习，也可以观看其他学生听取教师小组成员的建议。在去年举办第一届数据中心之后进行的一项调查发现，所有参与者（包括学生和教师）都认为数据中心非常有用。我们期望所有参加今年DC会议的学生（包括演讲者和刚刚参加会议的学生）都能对自己的工作重拾信心，对改进论文计划提出建议，并对什么是好的研究实践有更清晰的认识。

One of the most important aspects in learning how to do research is early and detailed discussions on various aspects of the research plan with experienced mentors. Unfortunately, most students get little feedback on their thesis at the initial stages except from their immediate advisors. Attending conferences can be a good way to meet other researchers and discuss work. However, the chances of a student at an early stage of his/her work getting the opportunity to discuss their preliminary ideas with well established researchers are not very good. The Doctoral Consortium (DC) at the American Association for Artificial Intelligence conference was planned as a way to improve this situation. Students about to begin their thesis work are chosen to present their research plans in front of a group of carefully selected faculty panelists. Unlike traditional conference presentations, the bulk of the time in a DC is devoted to discussions between the students and the faculty panelists. Each student is also assigned a faculty mentor who works with the student at a more detailed level. Students learn from both their own presentations as well as watching other students being advised by the faculty panelists. A survey taken after the first DC held last year found all participants (both students and faculty) agreeing that the DC was tremendously useful. We anticipate that all students attending the DC this year (both as presenters as well as those just attending) will leave with renewed confidence in their work, suggestions for improving their thesis plans and a clearer idea of what constitutes good research practice.