落园 » 经济学家们的社会网络(学术圈)|专注经济视角下的互联网

经济学家们的社会网络(学术圈)

时间过得真快,一写起论文来就是两三天没干别的,打开电脑开始coding...coding...整理数据、转换数据格式、调参数,好不热闹。可怜我的本本长期维持在马上就要烧主板的温度……这次的毕业论文写的很飘逸,固然与社会网络(social network)有关,但是技术上的东西都是借鉴的别人的,我想了半天加了一个贴现因子而已——其实模型真的很简单。为了让文章好看点,也长一点,内容也丰富一点,我就开始折腾着画图、画图,画完图又琢磨着总得用点计量啊,然后没事找事的把计算机模拟的结果拿到时间序列走了个过场——单位根检验。真想扁了自己,时间序列学的那么烂,唉。还得现翻计量书,丢人丢人。

论文大多数东西都是蛮无聊的,但是也有好玩的,那就是里面有几张经济学家们的社会网络的图——我乐于称之为“经济学学术圈”。这个idea是taiyun的,他做的是国内经济学家的,我就百无聊赖的又把国外的做了一遍,反正只消几十分钟数据就抓完了,何乐而不为?现在开始享受视觉盛宴吧!注:点大的代表其pagerank值高,其实就是联系多。

all_10_nolabel

经济学学术圈全貌(2000年后至今)

top_10_grey

经济学学术圈全貌,连接线用灰色隐去

all_10_main_all_label

经济学学术圈的结构,简化了很多点之后的效果

em_all_above3

计量经济学圈子,全貌

jpe_all

理论经济学圈子,全貌

jpe_main

理论经济学圈子的骨架

em_main

计量经济学圈子的骨架

更多图片请移步我的Flickr查看:http://www.flickr.com/photos/40637018@N06/sets/72157624032591252/

如果需要更多结果(比如更宽泛的时期、包含更多期刊数据等),请发邮件给我:cloudlychen@gmail.com

另,中文经济学学术圈子请移步taiyun的博客:http://taiyun.cos.name/2010/05/co-author-network-in-chinese-three-top-journal/

部分图像的PDF矢量版(可放大、查找)请移步我的英文博客下载:http://blog.cloudlychen.net/the-social-network-of-economists


Comments

  • Frank says:

    You might want to refine your graphs. In the econometrics picture, of course Peter C.B. Phillips and Phillips, Peter C.B., are the same guru.


    • cloudly says:

      Yes, you are right! In fact I'm looking for comments like yours. There must be many mistakes in these graphs. For example, "Jr." in theoretical world.
      I should admit that's the biggest problem facing by software. I don't know how to deal with this problem, since there are so many variations for a single name. Google Scholar considers "Coo Boo" as "C boo", so all person with same family name and the first letter of given name may regarded as one person. RePEc uses the similar way but they let the authors themselves to identify which are actually their works. It seems that maybe I can only fix them by hand...but there are thousands of records...


  • 世界草民 says:

    嗯,有意思,学习了!


  • 毛毛虫 says:

    😉 我也是学经济学的。呵呵。支持一下。


  • learner says:

    If you posted your source codes, maybe some people could tell you how to deal with your problems.


    • cloudly says:

      In fact, I use Gephi (from Gephi.org) to produce those graphs, so there is no source code.
      My problem is, how to identify the variations of name, like "Boo Aoo Coo" as the same of "A. B. Coo" or "B. A. Coo", etc... Could you give me some advice?


  • gaotao says:

    都喜欢上了社会网络呀..我们华中数模赛弄那个网站数据挖掘,也是画了个社会网络小图,炫一下评委的,嘿嘿


  • w says:

    一个简单的方法,从源数据中寻找同一个人不同名字的规则,几条规则应该可以覆盖80%以上的案例。然后用R对源数据进行处理即可。


    • cloudly says:

      恩,我现在想到的办法是把所有的middle name都省略,这样的话会减少很多重复的,那种“姓”和“名”互换的中国人我就没办法了……毕竟张莉和李章这种,不知道中文名的话没法区分……传说中生物学的英文杂志都开始允许中文署名了,雷啊~


  • causu says:

    请问,数据怎么抓呢?


Leave a Reply

Your email address will not be published. Required fields are marked *