大家来谈谈如何读paper吧~~

以前读paper的时候总是喜欢找下载到全文的...到现在做literature review还是这种习惯，但不幸的是前天被老板毙了...他说你总找原文干嘛...
牛牛们不要吝啬啊...comments，advices...

非常巧，最近教授在指导我们如何读paper，因为phd qualify的一项考试是给一篇paper，然后critique, 在3个教授面前做一个paper critique presentation，如果两个教授给通过，就通过了博士资格考试~

下面是我练习读paper事后，老师要我们考虑的几个问题，然后几个同学在一起集中讨论：

As part of your preparation for tomorrow’s discussion, I’d like each of you to do the following:

·         In a few sentences (2-3) describe the relevance of the described research area to society.

·         In a few sentences (2-3) describe what is novel about the work described in the paper.

·         List 3 strengths and 3 weaknesses of the paper. Make sure that you pick the right flavor of strengths/weaknesses, as prof. Sholl discussed last week; typos and obvious errors in equations don’t count.

·         Pick one of the key conclusions of the paper and analyze whether the data in the paper support that conclusion sufficiently.

·         Pick the one graph that you deem absolutely essential and be prepared to explain to the rest of the group within 90 seconds the key features of the graph AND their physical relevance. What is happening and WHY is it happening?

·         Find one relevant reference in the literature that is NOT listed in the paper; explain briefly why that reference is helpful for understanding the paper.

打印出来后，第一遍还是查单词吧，第二遍就是看作者提出的问题，怎么解决的，后面Appendix和技术细节倒不是最重要的。然后在paper的margin部分推推formula，写写comment什么的。有问题和导师约个时间问问什么的。我每篇paper看完两遍后上面都被涂得乱七八糟的。。。。。。。

然后paper后面的参考其实是非常重要的，很多时候同义一研究领域的都是串起来的。

看paper不看全文只看abstract吗？……哪能了解啥啊

先了解个大概，再细究。。。

我先看experimental section

看paper不看全文只看abstract吗？……哪能了解啥啊
afox0810 发表于 2010-10-2 06:31

至少４０％的paper 看完abstract就可以不用往下看了

6# rainfall1988

原来是这样。。

看看方法就好了。。然后把一些不同的地方整理出来。

反正也都是大同小异

我最近开始也只看abstract了

abstract+related work+ empirical study
我一般重点看这几个