下载掌阅APP，畅读海量书库
立即打开
I created the very first version of Gephi in Paris in 2007,while working in an institution called "Maison des Sciences de l'Homme" (the House of the Human Sciences).I was the only engineer of the lab and together with sociologists like Dana Diminescu we were facing difficult problems.We wanted to harvest specific areas of the World Wild Web,visualize them,and analyze them.But we could not find a suitable instrument,and so we decided to make one by ourselves.It was a little bit crazy but we were enthusiastic about this completely new challenge.
I drew inspiration from other works like Eytan Adar's GUESS and the well-known PAJEK software,but with a crucial difference:I focused on manipulability.I wanted to offer the possibility of"touching"networks,as if they were a physical thing.I believed that making networks handy was the only way to allow sociologists to explore our data.Dana accepted to play the role of "guinea pig" and I started to develop a prototype.Results were promising and we had the chance to get help from other engineers like Mathieu Bastian,who quickly became the most important person of the project:its lead developer.
Together we created much more than a simple tool.We created a complete open source software package with a sustainable architecture,a graphical user interface supporting different languages,and a multi-platform installer.We created a website allowing anyone to download it for free (gephi.org).We wrote academic papers and presented Gephi at international conferences.Eventually Gephi became used by scholars worldwide,in various disciplines,and has been downloaded more than a million times.Gephi is a living,breathing community and by holding this book in your hands,you are a part of it.
Gephi is now bigger than Dana,Mathieu,or me.None of us could have done it on his own,and we had to face the difficult transition from a small-scale challenge to an international project.Gephi's unexpected worldwide success was a source of joy as well as apprehension.How could you avoid disappointing so many people?It was both tiresome to fix long lists of issues,but also a relief to release new versions.We felt the strain of getting feedback from frustrated users,but also the excitement of discovering new plugins or tutorials from the community.Above all,we felt proud that so many people were willing to contribute to the Gephi source code,the website and the documentation,because it meant that they accepted to be part of the adventure.Gephi now also belongs to them.
For this reason I am honored to introduce you to this book,and I warmly thank Ooof for his efforts.He has won my utmost respect.This complete Gephi manual is the first of its kind in simplified Chinese,and will help you to understand how to use it from the basics to the most advanced features.It was written with the help of Yi Du,the prominent member of the community who developed the Twitter plugin (he deserves many thanks and recognition).A friend helped me to read it (since I do not read Chinese) and I realized that its 16 chapters cover each and every aspect of Gephi with detailed explanations and plenty of screenshots.Gephi is not as refined as commercial software,and it can be painful to use it if you ignore its many little problems and how to overcome them.You can learn Gephi by yourself but this book will make your life easier in addition to expanding your skills and knowledge in network analysis.Whether you are a beginner or an experienced user,I believe that you will find it an interesting read.
In this book you will find practical help such as an installation tutorial,a presentation of the interface,where to find plugins,examples of the different file formats (GEXF,CSV…),and even advanced features like dynamic networks and coding with the Gephi Toolkit.But more importantly,you will find a methodological assistance in the key steps of network analysis:why and how to use a layout algorithm,how to explore the network by filtering it,and how to properly use statistics like the PageRank or betweenness centrality.Gephi is not a "black box"providing you with a ready-made analysis,it is an exploration tool intended to support your own interpretative process.For this reason,it is very important to understand how the different features of Gephi help you making sense of your data.This is the most important mission fulfilled by this book.
Once you master the different aspects of Gephi,once you know how to delineate and compare the clusters of your network,once you can identify bridges and central nodes,once you are used to spend hours filtering your networks every possible way to understand their structure,then you will be able to "read" your networks as if they were regular maps.At this point,however,remember that we designed Gephi as an exploratory tool,and not as a map-making tool-despite including such features.Even if you understand an image of your network,most of the time others cannot because they do not have your knowledge.Gephi is more suited to manipulating your network in order to discover patterns than to build a clean and understandable image,which is often impossible.Gephi is like a microscope,except it looks into the complex rather than the miniscule.It is not a camera,it is a discovery tool.And as with any instrument designed to explore the world,your initial feeling may be a childish desire to play with it,purely for the experience it offers.In that case,know that I encourage you to do so!Playing is the first stage of getting familiar with a new technology.For all the following steps,you have this book.
I wish you a pleasant and instructive reading,and a lot of fun and exploration with Gephi.
Mathieu Jacomy
@jacomyma
2007年我在巴黎人文科学中心
工作的时候开发了Gephi的第一个版本。那时我是实验室里唯一的一名工程师，与我合作的是像Dana Diminescu
一样的社会学家。当时我们遇到了一个问题，我们想从互联网上收集一些特定领域的信息，并对它们进行分析和可视化。但是我们没有找到合适的工具来帮助我们完成这个任务。因此，我们决定自己开发这样一个工具。这听上去有些疯狂，但我们对这个新的挑战都充满热情！
我从Eytan Adar
开发的GUESS
和当时的知名软件PAJEK
那里得到了一些灵感，但是有一个非常关键的区别，那就是我们更关注可操作性。我想提供让使用者像触碰物体一样“触碰”网络的可能。我坚信让网络更容易操纵是让社会学家探索数据的唯一方式。Dana同意当我们的“小白鼠”，然后我着手开发原型系统。最终的结果非常令人满意，而且在开发原型的过程中，我们得到了其他工程师的帮助。其中一位就是Mathieu Bastian
，他也很快变成了这个项目的首席开发者。
我们在一起合作，开发出的不仅仅是一个小工具，而是一个完整的开源软件。这个软件具有易于使用的图形用户界面、支持多国语言、支持多种操作系统，且易于安装。我们制作了一个官方网站（gephi.org），还基于该软件撰写了一些学术文章，并在一些国际会议上展示了Gephi。最终，全球各个研究领域的科学家开始使用Gephi，目前已经超过了100万的下载量。最重要的是，Gephi仍然是一个活跃的社区，一旦你拥有了这本书，你就成为了这个社区的一员。
Gephi现在的发展远远超过了Dana、Mathieu或我的工作所达到的预期效果，我们中任何一个人都不能单独做到，目前还存在如何把一个小规模的项目扩展成一个国际级别项目的问题。Gephi在国际范围内意想不到的成功，既是一种幸福，也让我们忧虑，怎样做才能避免让大家失望呢？修改长长的问题列表里的问题，是一件很无聊的事情，但是最终发布新版本却是令人兴奋的。收到用户负面的反馈让人很失望，但是在社区里发现新的插件、教程又很让人兴奋。总而言之，我们骄傲人们愿意为Gephi贡献源代码、网站和文档，这意味着他们愿意和我们一起面对未来的挑战，Gephi属于他们！
正因如此，我很荣幸地向你介绍这本书，并且衷心地感谢ooof的努力，他赢得了我诚挚的敬意。这是一本填补了国内简体中文版Gephi专著空白的Gephi手册，它将帮助你理解如何使用Gephi的基本功能和高级功能。这本书也是在杜一的帮助下一起完成的，他开发了获取Twitter数据的插件，在社区里有突出的贡献，并且赢得了社区的尊敬。由于我不懂中文，我委托一个朋友帮我阅读了这本书，我发现这本书的16个章节包含了Gephi的方方面面，并且每一章都有详细的解释和丰富的图示。Gephi不如商业软件那样“精致”，所以你需要忍受它的一些小问题，并且想办法克服这些问题。当然，你可以自学Gephi，但是这本书会让你的学习过程更加容易，并且能够扩展你在网络分析上的能力。无论你是初学者还是有经验的用户，我相信你会从这本书中获益。
在这本书中，你会得到很多帮助，例如安装指南、界面组成、插件安装、各种不同类型文件的介绍（GEXF、CSV等）。除此之外，你甚至还可以学到很多Gephi的高级特性，例如动态网络和如何使用Gephi Toolkit。更为重要的是，这本书详细介绍了Gephi对网络分析的核心步骤的支持：如何使用布局算法、如何通过过滤网络来解释数据、如何合理的利用诸如PageRank或中心性等统计方法。Gephi对你来说不再是一个“黑盒子“，它是一个探索工具。因此一定要理解这个道理：Gephi不同的特性能帮助你更深入地掌握你的数据。这也是这本书最重要的目的。
一旦你掌握了Gephi的不同特性，一旦你懂得了如何描述和比较网络的聚类、一旦你能够发现网络中的桥接节点和中心节点、一旦你习惯了用几个小时的时间不断地过滤网络来深入理解网络的结构，你就可以像读普通的地图一样“阅读”网络，但一定要谨记：虽然Gephi本身有丰富的渲染功能，但Gephi设计的初衷是一个探索的工具，而不是一个普通的制图工具，即使你能够通过一张图就能理解一个网络，但是在多数情况下其他人由于没有和你相同的背景知识而无法理解，Gephi适合作为一个发现模式的工具，而不是作为一个制作漂亮图，形的工具，除了它能发现复杂性而不是微小的物体外，Gephi更像一个显微镜，而不是照相机。正如各种乐器一样，对于Gephi，你最初的感觉可能是像小孩一样想去把玩它，纯粹是因为它给你愉悦的体验，如果这样的话，我强烈推荐你继续下去！“把玩”是熟悉一个新技术的第一步，对于剩下的步骤，这本书可以帮助你。
希望你能从本书中获益，并且在使用 Gephi进行探索时获得满足。
Mathieu Jacomy
@jacomyma