Organizational  Chart Inference

组织结构图的推断

Nowadays, to  facilitate the communication and cooperation among employees, a new family of  online social networks has been adopted in many companies, which are called  the enterprise social networks" (ESNs). ESNs can provide employees with  various professional services to help them deal with daily work issues.  Meanwhile, employees in companies are usually organized into different  hierarchies according to the relative ranks of their positions. The company  internal management structure can be outlined with the organizational chart  visually, which is normally confidential to the public out of the privacy and  security concerns. In this paper, we want to study the IOC (Inference of  Organizational Chart) problem to identify company internal organizational  chart based on the heterogeneous online ESN launched in it. IOC is very  challenging to address as, to guarantee smooth operations, the internal  organizational charts of companies need to meet certain structural  requirements (about its depth and width). To solve the IOC problem, a novel  unsupervised method Create (ChArT REcovEr) is proposed in this paper, which  consists of 3 steps: (1) social stratification of ESN users into different  social classes, (2) supervision link inference from managers to subordinates,  and (3) consecutive social classes matching to prune the redundant supervision  links. Extensive experiments conducted on real-world online ESN dataset  demonstrate that Create can perform very well in addressing the IOC problem.

现在，为了方便沟通和运营，一种新系列的社交网络在很多公司被运用了，这种社交网络被称为“企业社交网络”（ESNs）。ESNs可以给员工提供很多专业的服务来帮助他们处理日常事务。同时，公司员工常常根据他们职位的相关排名被组织到不同的层级。公司内部的管理结构可以被这种组织结构可视化地勾勒出来，而这种组织结构常常由于隐私和安全关切是对公众保密的。在本文中，我们希望研究组织结构推断问题（IOC）来推断公司内部组织图，基于异构的在线ESN。IOC问题是一个非常具有挑战性的问题，为了保证平滑的运营，公司的内部组织结构图需要满足某种结构的要求（关于深度了宽度方面的）。为了解决IOC问题，本文提出一个新颖的非监督的方法（Create）。该方法包含3个步骤：（1）将ESN用户划分到不同的社交类别中（2）从管理者到追随者的监督链接的生成，以及（3）连续的社交层级的匹配来修剪冗余的监督链接。大量实际数据的实验表明，Create在处理IOC问题上表现非常好。

这是一篇典型的KDD的应用类型的文章。首先介绍问题背景（ESN），引出IOC问题。然后提出问题的Challenge，然后提出具体解决方法和步骤，最后说明一下实验。显然，这里的Contribution有以下三点（1）提出了IOC problem. (2)提出了解决的方法Create （3）大量实验验证了这个方法。